estimation, availability and production of tree biomass resources for energy purposes
TRANSCRIPT
Estimation, availability and production of tree biomass resources for energy purposes – a review of research challenges in Norway
Tron Eid Andreas Brunner Gunnhild Søgaard Rasmus Astrup Stein Tomter Øivind Løken Rune Eriksen
INA fagrapport 15
Department of Ecology and Natural Resource Management Norwegian University of Life Sciences
2010
ISSN 1891-2281
Estimation, availability and production of tree
biomass resources for energy purposes – a review of
research challenges in Norway
Tron Eid 1), Andreas Brunner 1), Gunnhild Søgaard 2), Rasmus Astrup 2), Stein Tomter 2), Øivind
Løken 1), Rune Eriksen 2)
1)
Department of Ecology and Natural Resource Management
Norwegian University of Life Sciences
P.O. Box 5003, NO-1432 Ås
2)
Norwegian Forest and Landscape Institute
P.O. Box 115, NO-1431 Ås
3
Contents Preface 5Summary 6Sammendrag 71. Introduction 82. Biomass estimation 10 2.1. Biomass equations and biomass expansion factors 11 2.1.1. Existing equations and their application 11 2.1.2. Research challenges 16 2.2. National forest inventory data in Norway 20 2.2.1. Background and present use of results 20 2.2.2. Sampling design 22 2.2.3. Volume and biomass estimation 23 2.2.4. Challenges in estimating biomass for energy purposes 26 2.3. Inventory methods for biomass estimation 28 2.3.1. Airborne and terrestrial LiDAR 29 2.3.2. Optical satellite imagery and satellite-based radar 30 2.3.3. Research challenges 313. Biomass resources and availability 33 3.1. Assessing present biomass and future potential 33 3.1.1. Previous studies 33 3.1.2. Research challenges 36 3.2. Environmental considerations 37 3.2.1. Background and previous research 37 3.2.2. Environmentally oriented instruments and their impact on forest production 38 3.2.3. Challenges in operationalizing the impacts 45 3.3. Economic considerations 50 3.3.1. Background 50 3.3.2. Biomass harvesting methods 51 3.3.3. Research challenges 54 3.4. Decision support systems 55 3.4.1. Existing decision support systems 56 3.4.2. Research and development challenges 584. Biomass production 61 4.1. Silvicultural methods in forestry 61 4.1.1. Density management 62 4.1.2. Use of pioneer species 65 4.1.3. Fertilization 70 4.2. Energy forests 71References 73
4
Preface The Bioenergy Innovation Centre (CenBio) is cooperation between the Norwegian University of
Life Sciences, Norwegian Institute of Forest and Landscape, Norwegian Centre for Agricultural
and Environmental Research, The Foundation for Scientific and Industrial Research (SINTEF),
and the Norwegian University of Science and Technology (NTNU). The project is funded by
The Research Council of Norway, a number of industrial partners and participating research
institutions.
The project started in year 2009 and will continue at least until 2013. The present report is a
review of the state of the art regarding the main issues that will be handled by Working Package
1.1 (Feedstock supply) in the CenBio project. The aim of the report is to indentify knowledge
gaps and challenges, and thus facilitate appropriate focus, development and priorities regarding
development and research within this subject area for the coming years.
Ås, 1th of November 2010
Tron Eid
5
Summary Eid, T., Brunner, A., Søgaard, G., Astrup, R., Tomter, S., Løken, Ø. & Eriksen, R. 2010.
Estimation, availability and production of tree biomass resources for energy purposes – a review
of research challenges in Norway [Estimering, tilgjengelighet og produksjon av biommasse i
skog for bioenergiformål – en gjennomgang av forskningsutfordringer i Norge]. INA fagrapport
15, 91 pp.
Ambitious targets for renewable energy production in Norway draw attention to biomass potent-
ials. The objective of this report is to review the state of the art regarding research on estimation
methods, the availability and production of tree biomass resources for energy purposes in
Norway in order to indentify knowledge gaps and thus facilitate appropriate focus, development
and priorities regarding research for the coming years. The review focuses on biomass from pri-
mary forest production with emphasis on Norwegian conditions, but also considers international
research, especially from the other Nordic countries. Three main subject areas are considered:
- biomass estimation
- biomass resources and availability
- biomass production.
The first part of this report comprises an overview of existing biomass equations and associated
inventory methods applied for estimating biomass in Norway. The overview includes a
description of the Norwegian National Forest Inventory data as a basis for large-scale biomass
assessments. The second part of the report comprises an overview of previous Norwegian
assessments of biomass as an energy supplier as well as suggestions for improvements in such
assessments. Improvement possibilities regarding the impacts of environmentally oriented
restrictions, appropriate models for productivity and cost calculations regarding biomass
harvesting systems, and implementation of biomass-related features in existing decisions support
systems to facilitate analyses, where timber production and biomass production for energy
purposes are equally important, are identified. The final part of the review focuses on silvi-
cultural options aiming at optimizing the value of total biomass instead of the conventional
approach to silviculture where the main focus is timber values.
Key words: forest biomass, energy, estimation, availability, production
6
Sammendrag Eid, T., Brunner, A., Søgaard, G., Astrup, R., Tomter, S., Løken, Ø. & Eriksen, R. 2010.
Estimation, availability and production of tree biomass resources for energy purposes – a review
of research challenges in Norway [Estimering, tilgjengelighet og produksjon av biommasse i
skog for bioenergiformål – en gjennomgang av forskningsutfordringer i Norge]. INA fagrapport
15, 91 pp.
Nye og ambisiøse mål for produksjon av fornybar energi i Norge setter fokus på biomasse i skog.
Hovedformålet med denne rapporten er å kartlegge kunnskapsstatus med hensyn på estimerings-
metoder, tilgjengelighet og produksjon av biomasse i skog i Norge. Denne kartleggingen er tenkt
å danne grunnlaget for en prioritering av utviklings- og forskningsoppgaver innen dette
fagområdet for de kommende årene. Rapporten legger vekt på norske forhold, men også
internasjonal forskning og utvikling, særlig fra de andre nordiske landene, blir tatt opp. Tre
hovedområder blir fokusert:
- biomasseestimering
- biomassetilgjengelighet
- biomasseproduksjon.
Den første delen omfatter en oversikt over eksisterende biomassefunksjoner og takstopplegg som
brukes i forbindelse kartlegging av biomasse i Norge. Oversikten inkluderer en beskrivelse av
Landsskogtakseringens opplegg og data som grunnlag for kartlegging av biomasse for større
områder. Den andre delen omfatter en oversikt over tidligere kartlegginger av tilgjengelig
biomasse i Norge, samt forslag til forbedringer som kan gjøres for slike analyser. Som viktige
områder for forskning og utvikling pekes det her på konsekvenser av miljørestriksjoner, modeller
for produktivitet og kostnader ved uttak av biomasse fra skog og tilrettelegging av eksisterende
prognoseverktøy slik at tømmerproduksjon og biomasseproduksjon til energi i større grad kan
sidestilles i analyser. Den siste delen av rapporten tar for seg skogskjøtselmetoder der en
fokuserer på verdien av total biomasse i stedet for den tradisjonelle tilnærmingen der en
fokuserer på verdien av tømmer.
Nøkkelord: biomasse fra skog, energi, estimering, tilgjengelighet, produksjon
7
1. Introduction New and ambitious targets for renewable energy production in Norway draw attention to
biomass potentials in the country. The Government has proposed to increase the annual use of
bioenergy by 14 TWh by year 2020 (St.meld. nr. 34 (2006-2007). Forest resources represent the
main biomass source for energy in Norway, partly as feedstock directly from forests, partly as
residues from the forest industry, and as a small portion of waste wood from, for example,
packaging goods and demolition of houses. In addition to forest biomass, biomass from
agricultural production and wet organic waste are important sources for renewable energy in
Norway.
The targets for future energy production from biomass have several challenging implications.
Suplies will have to be increased substantially compared to present use. Increased competition
for use of land and for biomass resources imply that present efficiencies must be improved.
Economic and ecological sustainability must be addressed throughout the value chain. To
determine how the quantities of biomass harvested for energy purposed may be doubled by the
year 2020 there is a need to implement research on several issues.
The main objective of the present report is to review the state of the art regarding research on
estimation, availability and production of tree biomass resources for energy purposes in Norway.
The aim is to indentify knowledge gaps and challenges, and thus facilitate appropriate focus,
development and priorities regarding research within these subject areas for the coming years.
The review is based on existing literature and ongoing research. The main focus is on Norwegian
conditions, but also international research and literature are considered.
Biomass derived from forest production applies to primary forest production (standing forest),
residues from the forest industry (secondary production), and waste wood from other activities.
This review focuses on biomass for energy from primary forest production. Based on these
priorities the following main subject areas are considered:
- biomass estimation
- biomass resources and availability
- biomass production.
8
Biomass estimation of a stand, forest property, or certain geographical unit or stratum is
generally done by multiplying biomass ha-1 by the total area under consideration. Biomass
equations and their application in Norway are described and discussed. Since the most important
source for biomass estimation at national and regional levels in Norway is National Forest
Inventory (NFI) data, the procedures applied to the Norwegian NFI are described, with emphasis
on biomass estimation. Also biomass resources related to areas normally not considered
appropriate for forest production (e.g. densely populated areas, pasture, power lines, roadsides,
border zones surrounding agriculture land) are discussed. Finally, the review examines inventory
methods and sampling designs relevant also for other purposes than national and regional
assessments, such as purposes related to practical management planning at different levels.
Assessments of biomass resources and the potential of biomass as a source of energy at national
and regional levels in Norway have been carried out by several projects over the past 10 years.
Most of these assessments have been relatively simple, and are therefore reviewed here in order
to identify knowledge gaps and suggest improvements. In addition to the basic biomass
estimation mentioned above, biomass resources are also considered with regard to
environmental, technical and economic restrictions. To assess the potential future availability of
biomass resources it is necessary to consider growth dynamics, temporal and spatial aspects,
harvesting costs, and biomass fraction prices. Forest management decision support systems (e.g.
a forest simulator and a decision module coupled with GIS functionality) already developed in
Norway comprise functionality that partly can deal with such aspects. There are obvious
shortcomings, however, for example related to new biomass harvesting methods and systems.
Previously developed decision support systems are described and possible measures for
improving them are discussed. The focus is on resources available at roadsides. Challenges
related to transport and logistics from roadsides to energy user are not considered. Questions
related to the market for biomass are only superficially considered in this review.
Biomass production may result from conventional forest production applying silvicultural
alternatives to optimize the value of timber. However, it may also result from conscious
management aiming to optimize the value of the total biomass. Such management could be
applied on sites previously used for conventional forest production but also for other areas, such
as those previously used for agricultural production. Accordingly, silvicultural aspects of
biomass production constitute the third main part of this review.
9
2. Biomass estimation Estimations of tree biomass quantities may be required for general forest planning or for studies
of energy and nutrients flows in ecosystems. Currently, however, the focus is on tree biomass
both as carbon stocks as a means to reduce atmospheric CO2 concentration and as an energy
source.
The total biomass of a given stand, forest property, geographical area, or stratum is generally
estimated by multiplying biomass ha-1 by the area under consideration. The basic requirement in
biomass estimation is the availability of appropriate species-specific biomass equations for
individual trees. A biomass equation may in general be used to estimate the dry weight of the
biomass for a whole tree or fractions of a tree by means of tree-level measured data (tree species,
diameter, height, etc.). The most common method for estimating biomass ha-1 is to take
measurements for all trees within the unit, thereafter apply biomass equations for individual
trees, and finally sum up biomass for all trees located in the unit.
If only a sample of trees within a given unit is measured there are principally three ways to
estimate biomass ha-1. First, the area unit can be characterized through stand-level variables (e.g.
species distribution, basal areal ha-1, mean height) and the biomass estimated by means of stand-
level biomass equations. Second, individual tree-level information can be generated by means of
diameter distribution models and/or height curves, and biomass can be estimated from equations
for individual trees. This method, however, requires available information on number of trees ha-
1. A third method is to apply stand-level volume equations and corresponding biomass expansion
factors (BEF) to estimate total biomass ha-1. A biomass expansion factor is a factor that converts
timber volumes into the dry weight of the biomass.
The choice of method for estimations of biomass ha-1 is closely related to the general inventory
design. In the Norwegian NFI or in procedures applied for airborne laser scanning, where all
diameters are measured within a sample plot, biomass equations for individual trees are applied.
For inventories based on ground-based laser scanning, photo interpretation, airborne radar
techniques, and satellite images it may be necessary to apply some of the other methods for
biomass estimation described above.
10
The first part of this chapter comprises a state-of-the-art description regarding biomass equations
and their application in Norway. Secondly, a description of procedures applied by the Norwegian
NFI, with emphasis on biomass estimation, is given. Finally, inventory methods, relevant for
smaller areas and practical planning, or for other purposes, are discussed.
2.1. Biomass equations and biomass expansion factors
2.1.1. Existing equations and their application
Extensive reviews of tree-level European volume and biomass equations have been given by
Zianis et al. (2005) and Muukkonen & Makipaa (2006). The total numbers of the compiled
equations for biomass estimation in these two reviews are 607 and 188, respectively. Many of
these biomass equations are developed for above-ground tree components only and most of them
are based on relatively few sampled sites, reflecting only limited variation regarding growing
conditions and silvicultural treatments, and a very limited number of sampled trees.
Most of the tree-level biomass equations developed in Norway are described in Table 1. In
general, they cover a relatively large number of species and fractions. The number of trees used
for developing models is also in most cases relatively high compared to the numbers reported by
Zianis et al. (2005) and Muukkonen & Makipaa (2006). However, the main problem with most
Norwegian equations is their general validity, i.e. most of them have been developed from data
covering limited variation regarding growth conditions and silvicultural treatments. For example,
the equations developed for birch (Opdahl 1987, Bollandsås et al. 2009) are based on data from
high altitude areas in southern Norway. This means that they should not be applied at low
altitudes or in other parts of the country, at least not before they have been tested for such sites.
Similar problems relate to the equations developed by Brække (1986) and Korsmo (1995), where
the data were based on sites subjected to certain silvicultural treatments, i.e. ditched peatlands
and regeneration by coppicing only, respectively.
The most widely applied biomass equations in Norway (Table 2) (see also section 2.2.3) have
been developed in Sweden (Marklund1987, 1988, Petersson & Ståhl 2006). Among the almost
800 biomass equations in Europe reviewed by Zianis et al. (2005) and Muukkonen & Makipaa
(2006), only Marklund’s (1987, 1988) equations were based on a sample size of several hundred
felled trees. Since these reviews, however, biomass equations have been developed in Finland
based on representative data material from several hundred trees for each of the three tree species
Norway spruce, Scots pine, and birch (Repola 2008, 2009).
11
In general, the trees used for the development of Marklund’s (1987, 1988) biomass equations
cover large variations regarding growing conditions and treatments in Sweden. The only
reservation made for the application of the equations was for treatments that ‘considerably
deviate from the normal’. When applied in Norway, the equations are definitely used outside
their geographical range. However, since they cover a wide range of conditions and treatments in
Sweden there is reason to believe that they will also be appropriate for use in Norway, at least in
southern and eastern Norway, and probably also in the central part of the country. Since wood
density and biophysical stem properties vary according to growing conditions, one should
probably be more careful when applying the equations in coastal regions in the western and
northern part of Norway without first checking their performance. Marklund’s equations for
birch are also frequently applied for other broadleaved tree species in Norway. This is clearly a
violation of the ‘data range’.
Wilhelmsen & Vestjordet (1974) developed stand-level biomass equations for Norway spruce in
Norway. These equations can be used to estimate biomass (dry weight) based on stand-level
variables such as basal area and mean height. However, since only the weight of the merchant-
able part of the stems can be estimated, the models have only limited value regarding biomass
estimation for energy purposes. No other stand-level biomass equations have been developed in
Norway.
The typical procedure when developing stand-level biomass equations, however, is to sum up the
biomass from all trees, estimated from individual-tree biomass equations, within a number of
stands, and thereafter regress the estimated biomass for the stands with different stand variables
(e.g. basal area and mean height). Example of such equations are given by Doruska et al. (2004),
used in the United States. Such an approach to the development of stand-level biomass equations
implies that the appropriateness and error structure of stand-level equations to a large extent rely
in turn on equations for individual trees. However, stand-level equations may be an appropriate
solution when no information on individual trees is available, such as in cases when forest
simulators based on area-based growth models are applied (see also section 3.4).
12
Tabl
e 1.
Bio
mas
s equ
atio
ns d
evel
oped
in N
orw
ay.
Tree
spec
ies
Uni
ts
No.
of
tre
es
Ran
ge o
f Fr
actio
ns*
R-s
quar
e
(a
ccor
ding
to fr
actio
ns)
Ref
eren
ce
B
iom
. D
H
D (c
m)
H (m
)
Nor
way
spru
ce (P
icea
abi
es)
g cm
-
35
2-15
-
AB
, BR
, DB
, FL,
SB
, SW
0.
99, 0
.98,
0.8
0, 0
.97,
0.
98, 0
.98
Bræ
kke
(198
6)
Sitk
a sp
ruce
(Pic
ea si
tche
nsis)
kg
cm
-
10
2-43
2.
6-34
A
B, B
R, D
B, F
L,
SB, S
W
0.97
, 0.7
3, 0
.51,
0.5
1,
0.99
, 0.9
7 Jo
hnse
n (2
009)
Scot
s pin
e (P
inus
sylv
estr
is)
g cm
-
16
2-15
-
AB
, BR
, CO
, DB
, FL
, SB
, SW
0.
99, 0
.97,
0.6
9, 0
.99,
0.
95, 0
.99,
0.9
9 B
rækk
e (1
986)
Scot
s pin
e (P
inus
sylv
estr
is)*
* g
cm
- 16
2-
20
- A
B, B
R, C
O, D
B,
FL, S
B, S
W
0.99
, 0.9
7, 0
.71,
0.9
3,
0.95
, 0.9
9, 0
.99
Bræ
kke
(198
6)
Birc
h (B
etul
a pu
besc
ens)
kg
cm
- 13
3 1-
8 -
AB
0.
97
Opd
ahl (
1987
)
Mou
ntai
n bi
rch
(Bet
ula
pube
scen
s sp
p. c
zere
panó
vii)
kg
mm
-
80
3-22
1.
8-12
A
B, C
R, S
T 0.
85, 0
.60,
0.9
1 B
olla
ndså
s et a
l.
(200
9)
Mou
ntai
n bi
rch
(Bet
ula
pube
scen
s Eh
rh, s
ubsp
. Tor
tuos
a Le
deb.
) kg
cm
cm
16
4-
12
3.7-
8 SW
, SB
, BR
, SR
0.
88, 0
.79,
0.9
2, 0
.94
Kje
lvik
(197
4)
Silv
er b
irch
(Bet
ula
pend
ula)
kg
cm
-
34-8
8 1-
13
1.5-
16
BR
, FL,
ST
0.92
, 0.7
5, 0
.99
Kor
smo
(199
5)
B
lack
ald
er (A
lnus
glu
tinos
a)
kg
cm
- 13
0-13
7 1-
14
1.5-
14
BR
, FL,
ST
0.86
, 0.7
7, 0
.98
Kor
smo
(199
5)
G
rey
alde
r (Al
nus i
ncan
a)
kg
cm
- 48
-54
1-15
1.
5-15
B
R, F
L, S
T 0.
89, 0
.84,
0.9
9 K
orsm
o (1
995)
Euro
pean
ash
(Fra
xinu
s exc
elsi
or)
kg
cm
- 18
-32
1-11
1.
5-14
B
R, F
L, S
T 0.
93, 0
.96,
0.9
8 K
orsm
o (1
995)
Euro
pean
asp
en (P
opul
us tr
emul
a)
kg
cm
- 70
-132
1-
15
1.5-
12
BR
, FL,
ST
0.92
, 0.9
4, 0
.99
Kor
smo
(199
5)
R
owan
(Sor
bus a
ucup
aria
) kg
cm
-
42-4
3 1-
10
1.5-
12
BR
, FL,
ST
0.90
, 0.8
1, 0
.99
Kor
smo
(199
5)
*A
B =
Tot
al a
bove
-gro
und
biom
ass,
BR
= B
ranc
h bi
omas
s, C
O =
Bio
mas
s of c
ones
, CR
= C
row
n bi
omas
s, D
B =
Bio
mas
s of d
ead
bran
ches
, FL
= To
tal f
olia
ge b
iom
ass,
SB
= B
iom
ass o
f ste
m b
ark,
SW
= S
tem
woo
d bi
omas
s, ST
= T
otal
stem
bio
mas
s (in
clud
ing
bark
), SR
=bio
mas
s of t
he st
ump
root
syst
em, *
* =
Ferti
lized
13
Tabl
e 2.
Bio
mas
s equ
atio
ns u
sual
ly a
pplie
d in
Nor
way
.
Tree
spec
ies
Uni
ts
No.
of
Ran
ge o
f Fr
actio
ns*
Func
tion
ID
R-s
quar
e R
efer
ence
B
io-
mas
s D
H
tre
es
D
(cm
) H
(m
)
(a
ccor
ding
to fr
actio
ns)
Nor
way
spru
ce
(P
icea
abi
es)
kg
cm
m
551
0-50
1-
30
CR
, DB
, FL,
RC
, RF,
SB
, SR
, ST,
SW
, SU
12
, 20,
16,
28,
31,
8,
23,
1, 5
, 26
0.99
, 0.8
7, 0
.97,
0.9
8, 0
.97,
0.
99, 0
.99,
0.9
9, 0
.99,
0.9
8
Mar
klun
d (1
987,
198
8)
Scot
s pin
e
(P
inus
sylv
estr
is)
kg
cm
m
493
0-48
1-
28
CR
, DB
, FL,
RC
, RF,
SB
, SR
, ST,
SW
, SU
14
, 22,
18,
31,
34,
10
, 25,
2, 6
, 28
0.98
, 0.8
8, 0
.95,
0.9
6, 0
.95,
0.
97, 0
.98,
0.9
9, 0
.99,
0.9
8 M
arkl
und
(198
7, 1
988)
Birc
h
(Bet
ula
pube
scen
s)
kg
cm
m
242
0-36
1-
24
CR
, DB
, SB
, ST,
SW
11
, 16,
8, 2
, 5
0.96
, 0.7
9, 0
.98,
0.9
9, 0
.99
Mar
klun
d (1
987,
198
8)
Nor
way
spru
ce
(P
icea
abi
es)
kg
mm
-
342
0-63
1-
35
SR
A-i
0.97
Pe
ters
son
& S
tåhl
(200
6)**
Scot
s pin
e
(P
inus
sylv
estr
is)
kg
mm
-
330
1-49
1-
27
SR
A-i
0.98
Pe
ters
son
& S
tåhl
(200
6)**
Birc
h
(Bet
ula
pube
scen
s)
kg
mm
-
14
1-27
2-
21
SR
A-i
0.95
Pe
ters
son
& S
tåhl
(200
6)**
Nor
way
spru
ce
(P
icea
abi
es)
kg
mm
-
342
0-63
1-
35
SR
B-i
0.97
Pe
ters
son
& S
tåhl
(200
6)**
*
Scot
s pin
e
(P
inus
sylv
estr
is)
kg
mm
-
330
1-49
1-
27
SR
B-i
0.98
Pe
ters
son
& S
tåhl
(200
6)**
*
Birc
h
(Bet
ula
pube
scen
s)
kg
mm
-
14
1-27
2-
21
SR
B-i
0.95
Pe
ters
son
& S
tåhl
(200
6)**
*
* C
R =
Cro
wn
biom
ass,
DB
= B
iom
ass o
f dea
d br
anch
es, F
L =
Tota
l fol
iage
bio
mas
s, R
C =
Bio
mas
s of c
oars
e ro
ots (
defin
ed in
the
study
), R
F =
Bio
mas
s of f
ine
root
s (d
efin
ed in
the
stud
y), S
B =
Bio
mas
s of s
tem
bar
k, S
R =
bio
mas
s of t
he st
ump
root
syst
em, S
T =
Tota
l ste
m b
iom
ass,
SU=
Stum
p bi
omas
s, SW
= S
tem
woo
d bi
omas
s, **
>5m
m, *
**>2
mm
14
�
A simpler method for estimating stand-level biomass was provided by Johansson (1999a, 1999b)
for Norway spruce and aspen growing on abandoned farmland. In this approach biomass equations
were first developed based on individual trees, and then the equations were applied based on
stand-level parameters (i.e. basal area mean diameter). Johansson (1999a) argued for such a
procedure based on the assumption that the variation in size between trees on abandoned farmland
where all trees have been planted was very low. Theoretically, a similar approach could be applied
also for other forest types in order to estimate stand biomass in cases where information on
individual trees does not exist. In Norway, information on basal area mean diameter (Dg), mean
height by basal area (HL), and number of trees ha-1 (N) is quite often available in forest
management plans and in area-based forest simulators. By applying these variables as input in, for
example, Marklund’s (1987, 1988) tree-level biomass equations, an estimate can be obtained for
biomass for a given stand. It is, however, obvious that such a procedure will cause biased
estimates at stand level due to variations in tree sizes. Even if the diameter distribution is narrow
and normally distributed, there will be a bias because of the non-linear relationship between
diameter and weight in the biomass equations. In most ‘normal’ stands considered as even-aged,
and particularly in uneven-aged forests, with a typical inverse J-shape in size distribution, care
should be taken in applying this approach as long as there have not been any systematic tests of
how large any biases will be that appear under different conditions.
As previously stated, biomass equations may be replaced by stand-level volume equations and
biomass expansion factors in order to estimate biomass. A biomass expansion factor (BEF)
converts timber volumes into the dry weight of biomass (see e.g. Lehtonen et al. 2004). Individual
tree volume and biomass equations are required as a basis to develop biomass expansion factors.
Since wood density and biophysical stem properties, among other characteristics, vary according
to the size and age of trees, it is necessary to establish relationships between the BEF and certain
stand characteristics. Lehtonen et al. (2004) in Finland and Jalkanen et al. (2005) in Sweden, for
example, established factors varying according to stand tree species and stand age. Although
applying BEFs is a good alternative for estimating stand-level biomass when information for
individual trees is not available, several authors have reported challenges related to this approach.
Jalkanen et al. (2005), for example, pointed to certain discrepancies relating to different
geographical regions and young age classes when testing the factors. Similar problems were
reported by Levy et al. (2004), who tested a number of biomass expansion factors for several
conifer species in Great Britain. Also, Albaugh et al. (2009) pointed to problems when applying
15
�
general (not site-specific) BEFs for sites in Sweden that had been subjected to certain treatments
(e.g. irrigation and fertilization).
2.1.2. Research challenges
Biomass estimation for carbon stock versus bioenergy
The main body of literature regarding biomass estimation over the past 10 years has focused on
biomass as carbon stocks, while less attention has been paid to biomass as an energy source. In
some cases, the two approaches coincide and equations can be applied irrespective of the intended
purpose. It is, however, important to be aware of differences between the approaches. In carbon
estimations the aim is to calculate the total biomass, while quantification of different biomass
fractions is less important. In estimations related to bioenergy it is important to quantify the
biomass that actually can be used for energy purposes. It is also important to distinguish between
different fractions since their potential for producing energy differs. In a study made by Palander
et al. (2009), for example, Marklund’s (1987, 1988) equations were tested at the operational level
of stump wood procurement and a calibrated biomass equation was developed to predict stump
biomass. The motivation for Palander’s study was to produce more accurate information for
planning and business management in energy production. Further examples of challenges related
to estimating biomass for energy production are described in the following:
- Norwegian national reporting on carbon stock (see section 2.2.3) is based on Marklund’s (1987,
1988) equations for all above ground biomass fractions, while Petersson & Ståhl’s (2006)
equations are applied for below ground biomass. Petersson & Ståhl (2006) found that their
equations gave 11% higher below ground biomass compared to Marklund’s (1987, 1988)
equations. The reason for this difference is the proportion of the roots included in the equations.
Petersson & Ståhl (2006) included all roots down to 2 mm, while Marklund (1987, 1988)
focused on biomass for bioenergy and, when excavating the stumps and roots, an unknown
proportion of roots was left in the ground. Although the latter approach probably is the most
realistic regarding biomass for energy, some conscious considerations on these differences
should be done.
- Marklund’s (1987, 1988) equations for birch did not include any equations for below-ground
biomass. Therefore, his equations for pine have been applied in Norway to estimate below-
ground biomass for birch. The fact that pine equations are applied implies an uncertainty that
should be considered.
16
�
- Marklund’s (1987, 1988) equations for stem and bark biomass also include the top. When
estimating forest residues (top and branches) after a conventional harvest for timber production,
the stem and bark proportion of the top have to be deducted from the stem and added to the
branch biomass. This proportion varies according to the stem taper and the grading and scaling
regulations for sawn logs and pulpwood (e.g. minimum top diameter required for pulpwood).
Proper analyses of the variation of this proportion based on Norwegian taper curves (Eide &
Langsæter 1954, Strand 1967, Blingsmo 1985) should therefore be carried out. Such analyses
should also consider other top-diameter requirements than those currently prevailing in order to
analyse variations in resources available for pulpwood and forest residues under different
assumptions for grading regulations.
- Several questions related to the proportion of needles, leaves and bark left in forests or at
roadsides and the proportion that actually can be utilized for energy purposes need to be
answered. Also the proportions of remaining wood in forests (short logs, small trees, etc.) have
to be considered in this context (see e.g. Gulbrandsen 1980, 1981, Asper 2007). Previous studies,
however, have covered only small areas and focused on traditional timber harvesting systems.
More research is therefore necessary to arrive at appropriate estimates according to harvesting
systems related to biomass for energy purposes.
Systematic evaluation of existing biomass equations for application in Norway
None of the Norwegian biomass equations are based on data covering a wide range of
geographical locations, conditions and treatments. Because of this, Marklund’s (1987, 1988)
equations, developed for Sweden, have been applied in Norway. Complemented with the
equations for below-ground biomass developed by Petersson & Ståhl (2006), there is reason to
believe that biomass is estimated in an appropriate way under most conditions and for the most
important tree species (Norway spruce, Scots pine and birch) also in Norway. Representative
biomass equations for the three most important tree species in Finland have been developed
(Repola 2008, 2009) and, in the same way as for the Swedish equations, these have potential for
covering most Norwegian conditions if applied.
The fact that comprehensive Swedish and Finnish biomass equations already exist implies that
large and expensive projects aiming at developing representative equations for Norway should not
be nesseary, at least not for spruce and pine. However, the performance of the Swedish and
17
�
Finnish equations under Norwegian conditions has never been systematically evaluated. Research
in Norway should therefore focus on a systematic evaluation of existing equations. The evaluation
should be approached in two different ways: qualitatively and based on empirical data. A
qualitative approach should include questions related to the representativeness and structure of the
data, the statistical methods used, the capability of the equations to describe the different
components of a tree, and the capability of applied independent variables to describe the biomass
of different fractions of a tree. The output from Marklund’s (1987, 1988) and Repola’s (2008,
2009) set of equations could be evaluated in relation to each other and compared with other
studies. The comparisons could be done for different conditions (e.g. productivity, treatments) and
geographical areas. A similar evaluation of Marklund’s equations has previously been done for
Finland by Kärkkäinen (2005).
An evaluation based on empirical data should also be carried out. Although no data exist that
facilitate systematic tests covering all conditions, there are several existing Norwegian datasets
from previous (see Table 1) and ongoing research that could be used for this purpose. In total,
these data sets could provide valuable input regarding the performance of the Swedish and Finnish
equations when applied in Norway.
The evaluation should, of course, cover as much as possible regarding the three tree species and
geographical areas. Since the coastal western and northern parts of Norway vary most from
Swedish and Finnish conditions, special attention should be paid to those areas. Also the
mountainous and high altitude areas in Norway should be paid attention in this respect. Norway
spruce, Scots pine and birch constitute approximately 92% of the growing stock in Norway
(Larsson & Hylen 2007). Most of the remaining volume is made up of different broadleaved tree
species. The fact that biomass equations for birch are applied also for other broadleaved tree
species requires special attention.
Biomass equations for birch in Norway
In Norway, downy birch (Betula pubescens Ehrh.) is the dominant broadleaved tree species from
sea level to the treeline. In addition, silver birch (Betula pendula Roth.) is abundant in eastern
parts of the country. Within the productive forest regions, birch makes up 14% of the country’s
total above-ground volume and approximately 25% of the forested area (Larsson & Hylen 2007).
Additionally, birch is the dominant species on approximately 2.5 million hectares of unproductive
18
�
forest. Historically, unproductive forest has not been covered by the national forest inventory and
neither volume nor biomass estimates are currently available.
It is well established that tree allometry and hence biomass fractions vary between tree species,
tree sizes, site types, stand types, and along environmental gradients. Consequently, to develop
biomass functions with reasonable precision and broad applicability, it is necessary to have data
from the same range of site type, stand type, geographic region, and environmental conditions
spanning a range of tree sizes (Marklund 1987). For example, a recent comparison between birch
below-ground biomass functions from Sweden and Finland showed differences in predicted values
of 15-20% (Repola et al. 2007). Some main objectives in a project related to biomass estimation
for birch could be to:
- Sample below- and above-ground biomass of individual trees of birch across a geographical and
environmental gradient.
- Characterize the error and variation associated with the currently applied biomass functions
through testing against the collected dataset.
- Develop new below-ground biomass functions that include environmental parameters as
predictor variables.
Biomass expansion factors for Norwegian conditions
As long as information on individual trees exists, tree biomass equations should be applied
directly and summarized to estimate stand-level information. In many cases, however, only stand-
level information is available, most commonly as volume ha-1, but also as basal area ha-1 and mean
height by basal area or as basal area mean diameter, mean height by basal area, and number of
trees ha-1. This applies to several management-oriented inventory methods as well as to analyses
required as input for decision support systems using area-based growth models (AVVIRK-2000
and Gaya-SGIS, see section 3.4.1). For all these cases the best option for estimating stand-level
biomass is probably to apply stand-level biomass expansion factors.
Biomass expansion factors for Norway spruce, Scots pine and birch under Norwegian conditions
could quite easily be developed from the combined tree and plot database of the Norwegian NFI
by applying volume equations (Braastad 1966, Brantseg 1967, Vestjordet 1967) and biomass
equations (Marklund 1987, 1988, Petersson & Ståhl 2006, Repola 2008, 2009) for individual trees.
19
�
Relationships (models) could then be established between plot-level (stand-level) information (e.g.
volume ha-1) and biomass expansion factor and biomass ha-1.
It is obvious that variations in biomass expansion factors according to tree species and stand age
need to be considered (see e.g. Lehtonen et al. 2004, Jalkanen et al. 2005). However, also forest
structure (e.g. even-aged versus uneven-aged), treatment history and regional differences should
also be considered. Further, special attention should be paid to young forests (development class
II) in this respect. Evaluations of such models to a large extent need to be based on qualitatively
based expectations regarding performance and previous research on, for example, variation in
wood density (e.g. Wilhelmsson 2001, Molteberg & Høibø 2007). However, as far as possible,
independent data from destructive sampling should also be applied in empirical testing.
2.2. National forest inventory data in Norway
2.2.1. Background and present use of results
The Norwegian National Forest Inventory (NFI) has been producing large-area forest resource
information since 1919. The first inventory was carried out in the period 1919–1930. The main
reason for the start of the inventories was concern about the forest situation and concern about the
possible lack of forest resources in future. To date, nine inventory cycles have been completed, of
which the most recent is for 2005–2009.
The inventory system and sampling design have changed considerably over the years. From 1986
to 1993, fixed-area circular permanent sample plots were installed in all counties except Finnmark.
The sixth NFI inventory (NFI6) was carried out on a county-by-county basis. In 1994 the concept
of continuous forest inventory was introduced, with 20% of the sample plots inventoried each
year. Since 1995, county-based inventories have been carried out together with reassessment of
the permanent plots in such a way that temporary plots have been measured in selected counties
and completed over a 5-year period. All counties (except Finnmark) have been covered during the
period 1995-2009. Limited inventory work in Finnmark County was started in 2005. The primary
aim was to provide data for carbon reporting (UNFCCC’s Kyoto Protocol), but it will also
contribute to more complete general forest statistics for Norway. This work is still ongoing (2010).
The field sampling intensity for permanent plots is basically the same throughout the entire
country, i.e. one sample plot represents the same forest area size regardless of the location of the
plot. There are, however, some exceptions to this principle. The number of temporary plots and
20
�
the distance between them has been adapted to the forest area in the county and to terrain
constraints.
The Norwegian NFI covers forests of all ownership groups, including protected forests and all
other land-use classes, and serves as a central forest information source for forest industries and
forest environment decisions and policy making. Information generated by the NFI has
traditionally been used for large-area forest management purposes such as harvest planning,
silviculture and forest improvement regimes at regional and national levels, and to some extent
decisions concerning development of forest industries. It also provides information to international
bodies such as the Food and Agriculture Organization of the United Nations (FAO) and its Global
Forest Resources Assessment (FRA) programme and Forest Europe (earlier the Ministerial
Conference on the Protection of Forests in Europe) (FAO 2007, MCPFE 2007). In addition, the
NFI provides information on forest health status and damage, biodiversity and carbon pools, and
changes for the Land Use, Land-Use Change and Forestry (LULUCF) reports of the United
Nations Framework Convention on Climate Change (UNFCCC). The Norwegian NFI has been
the main source of information for greenhouse gas reporting by the LULUCF sector to the
UNFCCC since the mid-1990s, and it will also play a central role in reporting LULUCF activities
under Articles 3.3 and 3.4 of the Kyoto Protocol.
Land-use categories reported by Norway are consistent with the IPCC Good Practice Guidance
(Penman et al. 2003). Area estimates by land-use categories, excluding some sub-categories of
cropland, are based on NFI data. IPCC land-use categories are formed on the basis of national land
classes using other variables assessed in the NFI. Data from the current inventory (2005–2009)
and three most recent NFIs have been used for estimating time series of areas for land-use
categories from 1990 to the present (2009).
Some new land-use related variables and classifications were included in NFI9 because data from
previous inventories did not provide adequate information about transitions between IPCC land-
use categories. These additions are expected to provide more accurate data for reporting emissions
and removals from afforestation, reforestation and deforestation (ARD) activities under the Kyoto
Protocol. All the LULUCF related carbon pool change estimates are based either directly or
indirectly on NFI data, with the exception of emissions ha-1 from organic soils (Rypdal et al.
2005).
21
�
In addition to conventional NFI variables such as stand and tree age, the following have been
introduced to the NFI: tree species composition, diameter distribution of the growing stock,
number of trees, soil and site variables, and variable groups describing forest biodiversity.
Assessments of the volume, diameter, tree species, decay class, and appearance class (categories
include broken snags, uprooted stems, broken lying stems, logging residues) of dead wood were
introduced for NFI7 (1994–1998) for all permanent plots. In subsequent inventories, only trees
that have died since the previous assessment have been recorded.
2.2.2. Sampling design
Norway is divided into 17 regions (counties) and for NFI7–NFI9 (1994–2009) each region has had
a different sampling intensity based on experiences from previous inventories, especially the total
land area of the county and the relationship between forested areas and total land area. The
regional sampling designs to some degree also reflect the variability in land-use classes and the
values of forest variables (growing stock by tree species). The basic observation unit is an
individual field plot. However, for practical reasons and for purposes of increasing the efficiency
of fieldwork, field plots are organized into clusters. The shape of the clusters may vary between
counties. The cluster design basically applies only to the temporary plots, while the permanent
plots are installed in a regular grid of 3x3 km in most parts of the country. An exception is
mountain areas (above the previously defined coniferous forest limit), where the spacing between
plots is 3x9 km. Multiple plot types were used in NFI9. The plot radius depends on the variable to
be recorded, and in some cases, also the value of the variable. The plots used in NFI9 are
described as follows:
1. For forests (productive and non-productive), other wooded land, and other land-use classes
where trees are assessed, the radius of the sample plot for tally trees is fixed at 8.92 m (area
250 m2). This plot type has been used for all permanent plots since 1994 for measuring trees
with dbh (diameter at breasth height) � 50 mm. Sample trees are now being selected using an
adjustable basal area factor that selects approximately 10 trees in each sample plot, when
possible.
2. Temporary plots are circular with an area of 250 m2 and include a smaller inner plot with an
area of 100 m2. All trees with dbh � 200 mm are measured in the larger plot, whereas only
trees with dbh � 50 mm are measured in the smaller, inner plot. In temporary plots, all trees
are considered tally trees, for which only species are observed and dbh is measured. No
22
�
sample trees are measured in temporary plots, because the same strata are represented by
permanent plots.
3. For all permanent plots with tree assessments, data for trees with dbh < 50 mm are collected
in four sub-plots of radius 1.3 m with centres located 5 m from the plot centre in directions
north, east, south, and west.
4. Circular sample plots of radius 17.84 m (1000 m2) centred at the centres of both permanent
and temporary plots are used to assess area-related data such as land-use class, crown cover,
development class, and site quality class. If a forest edge or stand boundary crosses the plot,
the plot shape may be altered to obtain the minimum area of 1000 m2 within the stand.
5. “Environmental Inventories in Forests” (manual for field work) also uses the NFI permanent
plots, but areas of 2000 m2 (radius 25.23 m) around plot centres are considered. The 2000 m2
plot should completely or partially overlap a specifically defined environmental unit (habitat)
in order for the plot to be used to assess the unit.
2.2.3. Volume and biomass estimation
Area estimates are based on the total land and inland water areas, which are known or assumed to
be error-free, and on the number of plot centres. In brief, the area estimate of a land stratum is the
product of the known land area and the ratio of the number of sample plots in the stratum and the
total number of plots. Because the number of plot centres on land is a random variable, area is
estimated using ratio estimators (Cochran 1977):
xyA
x
yAa n
1ii
n
1ii
s ��
�
�
�
� ,
where as is the area estimate of stratum s, A is the land area on the basis of the official statistics of
the Norwegian Mapping Authority, yi is 1 when the plot centre belongs to the stratum in question
and 0 otherwise, xi is 1 when the plot centre is on land, and n is the number of plot centres on land.
Some examples of land strata are forest land, spruce-dominated forest land, and forest land
thinned during the last five years.
The stem volume of trees in the Norwegian NFI is defined as the volume of the stem wood above
stump to the top of the tree. Volume estimation of individual trees from dbh and height is based on
traditional volume functions (Braastad 1966, Brantseg 1967, Vestjordet 1967, Bauger 1995). The
functions are available for individual tree species (spruce, pine, birch) and also to some degree for
23
�
various regions (i.e. separate functions for western Norway exist). Both volume under bark and
over bark are calculated and stored in the database, but for national purposes volume under bark is
commonly used. Growing stock includes stem volumes of all living trees, independently of
species. For national reporting only the volume of trees with dbh � 50 mm is usually included, but
from NFI9 it will be possible to report volume for all trees with dbh > 0 and a minimum height of
1.3m. The estimates of dead wood are intended mainly for biodiversity assessment, but will also
be of importance for estimations of carbon pools. A complete assessment of dead wood was
carried out 1994–1998. Subsequently, only the trees that have died during the periods between two
assessments of the permanent plots have been recorded.
Volumes of individual trees and volumes by various strata are predicted for sample trees
(approximately 10 in each sample plot) using volume functions based on dbh and h. These volume
estimates are used directly in the following calculations. From the sample trees, an average tariff is
calculated individually for tree species classes ‘spruce’, ‘pine’ and ‘broadleaved trees’ occurring
in the sample plot. Assuming a reasonably constant diameter-height relationship on the plot,
volume can be estimated using only dbh. The volumes of tally trees are predicted by sample plots,
applying the tariffs from the sample trees:
1. Volume ha-1 and total volume represented by each sample plot are calculated from volume per
tree (and plot), sample plot size and area represented by each plot.
2. Mean volumes are tabulated by computation strata.
3. Total volumes are tabulated by computation strata.
The volume estimators are ratio estimators similar to the area estimators. Briefly, to obtain the
mean volume for a given stratum, the mean volumes of all trees belonging to that stratum are
added and the total is divided by the number of field plots in the stratum. The mean volume of a
tree for this purpose is volume ha-1 represented by the tree. The indicator variable yi in the
numerator of (1) is replaced with the volume ha-1 represented by a tree, or the volume ha-1 of
timber class of interest represented by the tree, on field plot i when computing volume ha-1 or total
volume estimates. For total volumes, the volumes ha-1 must be multiplied by the area estimate for
the stratum in question. This operation has normally been performed for each sample plot and is
available in the NFI database; hence the total volume can be estimated by summing up the values
for all plots in the stratum.
24
�
The estimation procedure for biomass refers to all living trees with a height of at least 1.3 m.
Thus, small trees, shrubs and other vegetation, such as herbs are not included in the figures. The
biomass of trees with a stem diameter larger than 50 mm measured 1.3 m above the ground is
individually monitored.
Both above-ground biomass and below-ground biomass are reported. Above-ground biomass is
defined as living biomass above stump height (1% of the tree height). The Swedish biomass
equations for individual trees developed by Marklund (1987, 1988) are applied for predicting the
various tree biomass components: stem, stem bark, living branches, dead branches, and needles
(not leaves) of Norway spruce, Scots pine and birch. These species (including approximately1%
other coniferous species than Norway spruce and Scots pine) constitute 92% of the growing stock
(Larsson et al. 2007). Other broadleaved species than birch constitute the remaining 8%, and the
birch functions are applied to all these broadleaved species. Below-ground biomass is defined as
living biomass below stump height down to a root diameter of 2 mm and is estimated using
biomass equations (Petersson & Ståhl 2006) for the same tree species as for above-ground living
biomass. The living biomass is estimated consistently based on the same monitoring design, by
using the same functions for the same tree species from the base year 1990 and onward.
The biomass for all trees larger than 50 mm in dbh was estimated from their diameter and height
assessments. Trees with a diameter less than 50 mm may be included in estimates after 2009,
when a full cycle of assessments of small trees has been completed.
NFI6 (1986–1993) was the last inventory where fieldwork proceeded by regions. Since 1994, the
inventory has been performed continuously with all permanent plots reassessed every 5 years. For
example, the reference year for data collected over the period 2001–2005 is 2003, and for the
period 2002–2006 it is 2004. By using such a moving average, both volume and biomass changes
have been estimated. The reference year moves forward one year when data for a new season is
included in the calculations and data for the oldest season is excluded.
Error estimates are usually based on formulas traditionally used for random sampling, under the
assumption that they provide an adequate approximation to the errors for systematic sampling. For
the sampling error of area estimates, the normal approximation to the binomial distribution
(Snedecor & Cochran 1989) is used, and area standard errors, and volume/biomass standard errors
are combined in order to obtain the standard error of the total volume or biomass estimates.
25
�
2.2.4. Challenges in estimating biomass for energy purposes
There is no other source better suited for providing large area estimates of biomass for energy
purposes than the NFI data (see section 2.3 for alternative inventory methods). These data are
based on a well-documented sampling design and on detailed fieldwork instructions. The sample
plots cover most of the land area of Norway in a systematic grid providing representativeness in
terms of the estimates, and many variables are recorded for each plot describing forest and site
conditions. These data therefore provide a solid basis for estimating current biomass resources as
well as for assessing future potential.
For biomass assessments related to bioenergy it may be necessary to quantify the biomass
according to different strata. Examples of such strata are geographical units, land cover and land-
use categories, forest characteristics, site characteristics, and biomass fractions. Some challenges
when quantifying biomass according to these strata are discussed in the following:
- Geographical units
The Norwegian NFI regularly reports on growing stock and biomass resources for the country as
a whole, for regions (i.e. several counties), and for individual counties. Results for smaller
geographical units than a county are usually not reported because the requirements for maximum
standard errors may be violated due to low numbers of sample plots.
The size of the land area, and accordingly also the number of sample plots distributed, varies
considerably between counties. In Hedmark County, for example, the total number of permanent
sample plots is approximately 2400, while in Vestfold County the number is approximately 200.
Stratification within a county according to, for example, land cover and land-use categories, or
according to forest and site characteristics therefore has to be considered carefully in each
individual case. For strata with few sample plots (e.g. certain land-use categories) volume and
biomass estimates can only be presented at a regional level.
Since all plots are georeferenced, it is possible to do specific analyses by for example estimating
the biomass resources present within certain distances from an anticipated location of an energy
producer. In such cases it is important to consider especially the number of sample plots within
the area, and the corresponding standard error for the estimated biomass.
26
�
- Land cover and land-use categories
The public debate on bioenergy resources in Norway is quite often related to high expectations
regarding marginal areas (e.g. pasture, power lines, roadsides, border zones surrounding
agriculture land). There is a lack of proper definitions and descriptions for such areas, and there
is also confusion related to whether or not such areas are included in NFI assessments.
Over several years, the NFI has covered forest and other wooded land below the coniferous
forest limit. However, inventorying sub-alpine forests and other wooded land was not started
until 2005, and the first cycle was completed in 2009. An inventory of Finnmark County was not
started until recently, and so far only the coniferous forests have been completed and the work is
still ongoing in the broadleaved forests.
For the land cover and land-use classes ‘productive forest’, ‘non-productive forest’ and ‘other
wooded land’, where land use is classified as ‘forestry’, ‘forest reserves’ and ‘outdoor recreation
area’, all living trees with dbh � 5 cm are measured. In addition, assessments are done for
smaller trees (on sub-plots) and for dead standing trees. A complete assessment of dead lying
trees has not been carried out since the mid-1990s, but are introduced again in 2010. Currently,
only a rough estimate can be made for lying trees. However, this means that practically all
biomass can be estimated for all these land cover and land-use categories.
Some areas outside forests and other wooded land are surveyed or partly surveyed by the NFI.
Thus, an estimate of the potential of these areas for biomass supply may be obtained from
existing data. To these belong areas designated for holiday cabins, pastures, Calluna heath (near
the western coast), power lines, and gardens. Still absent from the NFI inventories are trees in
built-up areas, military areas for weapons training and shooting, roads, railways, and airports.
Some of these areas are not accessible, neither for inventory, nor for utilization of wood and
biomass resources. It may be possible to provide some rough estimates based on their extent,
together with an estimated growing stock and biomass ha-1, but the significance of such areas
should be expected to be fairly low.
- Forest characteristics
For several reasons, information on quantities of biomass resources according to different forest
conditions (e.g. tree species, development class, density) may be interesting from an energy point
of view. Tree species, for example, are directly related to the efficiency of energy production,
27
�
density (biomass ha-1) is related to harvesting costs, and development classes are related to
availability of resources. Sample plots may be stratified according to such variables. However,
also here the main challenge is to avoid violating the requirements for maximum standard errors.
- Site characteristics
Many variables characterizing the growing site conditions for sample plots are directly recorded
in the field or are indirectly available through databases or digital maps. Examples are site
quality, altitude, latitude, longitude, soil and rock properties, and climate variables like
precipitation and temperature. A link between these variables and the associated biomass
quantities and qualities may provide interesting data related to different uses of the biomass.
Variations in, for example, moisture content, wood density, wood structure, and ash content
according to growing site characteristics may have a practical impact on fuel quality and heat
value or on potential use of residues, i.e. ashes, for fertilizing (see e.g. Sikström et al. 2009).
Here, too, the main challenge is to avoid violating the requirements for maximum standard error
when stratifying NFI data.
- Biomass fractions
When considering carbon sequestration, all above- and below-ground biomass should be in-
cluded. For bioenergy it is important to quantify the biomass that actually can be used for
energy. All roots, for example, can hardly be expected to be used for energy. It is also important
to distinguish between fractions because they have different potential in producing energy. There
is also uncertainty related to the proportion of needles, leaves and bark that can be utilized for
energy purposes, and to the proportions of waste left when harvesting biomass.
In general, for the permanent sample plots, all trees with dbh � 5 cm are measured for diameter.
However, smaller trees are also measured for sub-plots (see section 2.2.3), and biomass can be
estimated. Also biomass of dead standing and fallen trees may be estimated. For both these
categories, smaller trees and dead trees, there are several challenges regarding the proportions
that can be utilized for energy purposes.
2.3. Inventory methods for biomass estimation
There are several available approaches to biomass estimation that vary in cost, spatial resolution,
and associated error levels. The appropriateness of a given forest biomass inventory method will
depend on the intended use and associated willingness to pay and accept errors. The historically
most common approach for biomass estimation is to combine field plots with biomass equations
28
�
or biomass expansion factors (see section 2.1). For sampling-based biomass estimates for large
regions or nations, such as provided by the Norwegian NFI (see section 2.2), field measurements
combined with biomass equations is the approach applied in all Scandinavian countries. On the
other hand, for applications where a full census of stand-level biomass is required, utilization of
field-based methods may often be associated with high costs. Thus, remote sensing based
approaches are often preferred for inventories associated with strategic and operational planning
of forest estates. In the following sections we provide a brief overview of different inventory
methods and their applications and future research possibilities.
All remote sensing based methods mentioned in this report are dependent on field plots for
calibration and validation. A remote sensing method will never be better than the field data used
for calibration. Thus, to obtain good results with remote sensing methods high quality field plots
in combination with good biomass equations are required.
2.3.1. Airborne and terrestrial LiDAR
Airborne LiDAR (Light Detection and Ranging) as an inventory tool has seen large development
in the past 10 years (see reviews in Næsset 2007 and Næsset & Gobakken 2008) and is now an
operational tool used for most forest planning in Norway (Næsset 2007). Biomass estimates can
either be obtained from LiDAR data by single tree segmentation (e.g. Solberg et al. 2006) or by
the area-based method (e.g. Næsset 2007). In single tree segmentation, individual crowns are
detected and variables such as tree volume and biomass are regressed based on tree height and
crown characteristics extracted from the laser data. In the area-based method, plot-level variables
such as biomass and volume are regressed directly from plot-level height percentiles in the laser
data. Single tree segmentation requires laser data with more laser pulses per area than the area-
based method and single tree segmentation is hence associated with substantially higher costs.
The stand-level error associated with an airborne LiDAR-based biomass estimate will vary
between individual projects and depend on the actual LiDAR acquisition, the quality and quantity
of data for calibration, and the applied algorithms. Still, when implemented in a reasonable way,
LiDAR-based stand-level volume and biomass estimates have been reported with mean squared
errors of about 10-14%, which compares favourably with most other methods (Næsset 2007).
In Norway, the cost of airborne LiDAR-based inventory compares favourably with field-based
approaches. Still, large regions wall-to-wall coverage of LiDAR data is associated with very high
29
�
cost. Thus, this technique may only be applied in limited areas, or in sampling-based schemes for
routine inventories of large areas such as countries (Næsset et al. 2009). The high cost has given
rise to research into sampling-based inventorying with LiDAR (Hedmark prosjektet, UMB) and
lower cost satellite-based approaches.
Terrestrial LiDAR has received less attention in Norway than airborne LiDAR. While airborne
laser scanning has the potential to cover large areas at a reasonable cost, terrestrial LiDAR has the
potential for giving detailed estimates for small areas. Internationally, terrestrial LiDAR has been
used to measure tree diameters (e.g. Henning & Radtke 2006a), crown attributes (e.g. Henning &
Radtke 2006b), and responses to silvicultural treatments (e.g. Huang et al. 2007).
Recently, research activities related to terrestrial LiDAR has commenced in Norway. The
BALABU (Bakkemontert laser som verktøy for bedre utnyttelse av skogressursene) project
(owned by Viken skog) is investigating inventory methodology with terrestrial LiDAR and how
terrestrial LiDAR can be utilized in combination with airborne LiDAR. The BALABU project is
carried out in collaboration with the Irish company Treemetrics which bases its business on
carrying out forest inventory with terrestrial LiDAR.
2.3.2. Optical satellite imagery and satellite-based radar
There is a wide range of remote sensing techniques that utilize optical imagery from satellites to
create wall-to-wall maps of forest conditions and resources (e.g. Luther et al. 2006, Gjertsen 2007,
Olander et al. 2008, Tomppo et al. 2008). The advantages of optical satellite imagery are
competitive prices (many data sources, such as Landsat TM imagery, are free), coverage of large
areas, and repeated coverage. The disadvantages are problems with cloud coverage (see e.g.
Olander et al. 2008), problems with distinguishing trees from other vegetation, and a saturation
effect for biomass that makes it hard to distinguish biomass concentrations above a certain
threshold (e.g. Nilson et al. 2003, Eriksson et al. 2006).
In Scandinavia, the most developed application of satellite imagery for forest recourses is Multi-
Source National Forest Inventory (MSFI) (e.g. see review of method and applications in Tomppo
et al. 2008). MSFI was developed in Finland to provide forest statistics for small areas that are
equivalent in quality to the regional or national estimates provided by national forest inventories
(Tomppo et al. 2008). In MSFI, plot data are combined with satellite images (Landsat) and the K-
Nearest Neighbor estimation method. The product of MSFI is statistics for small areas and wall-
30
�
to-wall forest maps. Due to the estimation method, all variables included in the field plots are
estimated in the MSFI. The root mean squared error of the MSFI is substantial (> 12%) for small
areas (< 100 ha) but becomes smaller as the area of interest increases (Tomppo et al. 2008). The
Norwegian MSFI product (Gjertsen 2007) is named SAT-SKOG, and maps with coverage for
southern and eastern Norway are currently available on the Internet (www.skogoglandskap.no).
The use of radar as a tool in forest monitoring is currently an international research topic
associated with high expectations (e.g. Olander et al. 2008). Synthetic Aperture Radar (SAR) is an
active sensor that ”sees” through clouds and works in any weather. Data can hence be obtained
more reliable in cloudy regions such as the tropics or parts of Norway. Recently, several types of
radar data from both airborne and satellite platforms have been investigated for mapping forest
attributes (e.g. Magnusson & Fransson 2004, Eriksson et al. 2007, Sexton et al. 2009). In Norway,
Weydahl et al. (2007) utilized SRTM X-band InSAR to derive vegetation heights that
corresponded to between one-half and two-thirds of the tree heights. Simultaneously, initial
studies have shown that x-band InSAR has potential for mapping forest biomass (Solberg et al.
2010a, 2010b) with a relatively low level of associated errors. A newly founded study (Research
Council of Norway project led by Solberg and Astrup) will further investigate the utility of radar
in forest inventorying in Vestfold County.
2.3.3. Research challenges
The main research challenge for biomass estimation is to improve current methods or create new
and better methods. Such improvements can result in better coverage of biomass inventories,
decreases in errors associated with existing methods, and cost reductions relating to methods.
Selected research opportunities could include the following:
- Terrestrial LiDAR and biomass estimation. Given that terrestrial LiDAR can be utilized to
estimate detailed crown attributes it is very likely that the technology can be utilized directly to
estimate biomass in crowns and stems without utilization of biomass functions or biomass
expansion factors. Given the uncertainty associated with existing biomass functions and biomass
expansion factors, a direct field measurement tool may be very useful for field-based inventories,
in research plots, and for calibration plots for airborne remote sensing techniques.
- Radar and biomass mapping. Given the initial Norwegian results from biomass estimation with
radar, further research may result in a cost-efficient method for biomass mapping of large areas.
31
�
Satellite-based InSAR data can be provided for the whole of Norway at a relatively low cost and
can potentially be utilized to provide biomass estimates with a relatively low error level.
- Improvement of MSFI and use of maps. The SAT-SKOG product is the only biomass map that
provides a uniform coverage of a substantial part of Norway. Utilization of SAT-SKOG to
analyse biomass availability in different regions and in relation to bioenergy production facilities
may be of interest to the industry. Improvement of SAT-SKOG with error estimates and
extension of SAT-SKOG to cover all of Norway may also be a topic for development.
- Airborne LiDAR. Airborne LiDAR and forest resource mapping has been the topic of much
research in recent years and much research is currently in progress. Given the wide operational
application of this technique, research that leads to improved LiDAR-based estimates or reduce
costs may be very beneficial for the forestry sector in Norway.
32
�
3. Biomass resources and availability 3.1. Assessing present resources and future potential
3.1.1. Previous studies
Over the past 10 years several national studies related to potential use of biomass for energy
production in Norway have been reported (OED 1997, NOU 1998, KanEnergi 2003, 2007). These
studies have several aspects in common. In addition to assessment of the potential from primary
forest production (standing forest), they have also considered biomass-related residues from the
forest industry, and partly also biomass from agricultural production of plants and biomass from
wet organic waste and other sources. The studies have also compared the present use of biomass
for energy purposes with the assessments they have made on the potentials. The main conclusion
has been that there is a large gap between use and potential, i.e. technically it is possible to
increase biomass production for energy purposes considerably, especially from primary forest
production. However, a further conclusion made in the studies is that such an increase hardly will
be seen without changes in external conditions (prices, costs, subsidies, taxes, prices for
alternative products, etc.).
The analyses of the studies on the assessment of potential biomass from forest production have
been relatively simple. In general, the starting point has been national (in some cases regional)
estimates of total annual growth. From these estimates of total growth, a net annual potential for
biomass production has been estimated based on quite rough assumptions related to environmental
considerations, present harvesing methods and levels, and residues from this harvesting activity.
To some extent harvesting costs and prices for different biomass qualities and quantities have been
considered.
Applying annual growth estimates as a starting point in the assessments of potential biomass
production is a rather static approach. Changes over time in harvest and growth levels, and relative
changes between harvest and growth levels, and the subsequent long-term consequences of such
changes, can hardly be considered appropriately in this respect. A more dynamic approach is
necessary to capture the changes. Using aggregated growth estimates at national level also means
that no real spatial considerations are done. To capture temporal as well as spatial aspects, a
decision support system with a growth simulator describing the forest dynamics is required.
33
�
A more comprehensive approach to assessing the potential of biomass production for energy
purposes has been done by Langerud et al. (2007). In their study, they applied the AVVIRK-2000
model to describe potential future forest resources and harvests. The present state of Norwegian
forests was described by means of 8700 sample plots from the NFI covering all productive forests
with area utilization ‘forestry’ in Norway. To assess the biomass potential they first estimated the
long-term maximum sustained yield. Secondly, they reduced this quantity based on environmental
and economical (harvesting costs) considerations to a net maximum sustained yield. This was
finally compared to the actual harvest level observed over the past 10 years, and the difference
between the net maximum sustained yield and the current harvest level was regarded as the
potential for increased production of biomass for energy purposes. Similar approaches for
different regional levels with the application of AVVIRK-2000 have been described by
Hobbelstad (2007a, 2007b). In the most recent national assessments carried out according to the
methods described by Langerud et al. (2007), the annual potential harvest of biomass for energy
purposes in Norway was estimated to be equivalent to 26.8 TWh (Gjølsjø & Hobbelstad (2009).
Rørstad et al. (2010) provided a new dimension to biomass availability analyses in Norway when
they developed cost-supply curves for a region. In their study, firstly Gaya-SGIS was applied to
optimize conventional timber harvesting level alternatives according to net present value, and
secondly cost-supply curves for forest residues following as a consequence of the conventional
timber harvesting were developed by means of productivity models and costs related to the
harvesting and transport of residues. The analyses comprised spatial as well as temporal aspects
since stand-level data from 40,000 hectares of forest were considered for a period of 50 years. The
main improvement achieved in the project was the addition of cost-functions describing variation
in residues, harvesting costs according to forest conditions, and geographic dispersion.
In order to assess the future potential utilization or availability of biomass resources it is important
to consider growth dynamics, temporal and spatial dimensions, environmentally oriented
restrictions, and harvest costs and prices, and also to optimize the linkages between all these
aspects. To some extent, Gaya-SGIS comprises functionality to deal with this. Previously,
comprehensive analyses have been done applying this model at national and regional level based
on data from the Norwegian NFI (Hoen et al. 2001, Eid et al. 2002). These studies focused on
optimization, cost-efficiency and potential supply according to conventional timber production.
No considerations, however, were made of biomass supply for energy purposes, i.e. the concepts,
methods and assumptions made were directed towards timber production (see Hoen et al. 1998).
34
�
The body of international literature on assessments of biomass resources, availability and supply
potential is vast. The approaches and tools applied for such analyses cover a wide range regarding
comprehensiveness and sophistication. Assessments of biomass potential supply have been
provided even at global and European level (e.g. Berndes et al. 2003, Karjalainen et al. 2004,
Ladanai & Vinterbäck 2009). In the following, a few examples from neighbouring countries to
Norway are described.
The Swedish Forest Agency (2008) in Sweden has provided extensive long-term scenarios
assessing potential timber and fuel biomass resources by means of NFI sample plots and the
HUGIN decision support system. The main aim of the analyses was to study the timber production
potential, and therefore the bioenergy potential was assessed according to a reference scenario
based on present ambitions for timber production and present environmentally oriented legislation.
The biomass quantities assessed were branches and tops from final felling, stumps from final
felling, branches and tops from thinning, stumps from thinning, and total above-ground biomass
from young growth tending. In addition, to the general environmentally oriented considerations
certain assumptions were made regarding biomass to be left in the forest to secure sufficient
nutrient supply in the long run. The HUGIN analyses were also followed up by a study where
marginal cost curves (accumulated biomass according to increasing harvesting cost) at national
level were constructed for forest residues and stumps for a given final harvest level (Athanassiadis
et al. 2009).
An interesting study from Denmark was provided by Nord-Larsen & Talbot (2004), where they
estimated potential fuel wood resources based on NFI data. In the first stage they estimated the
total potential of biomass resources. In the second stage the economically available resources were
quantified by coupling biomass data from NFI sample plots data with location of conversion
facilities and projected consumption. The data were incorporated into an economic model based
on GIS that formed industrial marginal cost-of-supply curves from an optimization (linear
programming) of the allocation of fuel wood. The main conclusion of the assessment was that
resources in Denmark were able to meet 57–100% of the projected future consumption.
Padari et al. (2009) has estimated Estonian wood fuel resources in a study based on field
measurements (not NFI data), digital national map layers (base map, soil map, forest cover, etc.),
models for growth, and a number of assumptions related to rotation ages for different site classes
35
�
and tree species. Also different environmental restrictions and harvesting systems were
considered. Theoretical long-term average annual yield estimates were presented for merchantable
timber resources as well as energy resources from harvesting residues and stumps. Attempts to
provide assessments of forest energy resources at national level have also been provided in
Lithuania by Kairi�kštis & Jaskelevi�ius (2003). In Finland and Sweden many studies on biomass
availability have been carried out focusing on developing new methods as well as providing
updated quantifications (e.g. Malinen et al. 2001, Bååth et al. 2002, Ranta 2005).
When discussing potential biomass availability it is, of course, also important to consider the
interaction between forestry and forest industry. The development of forest-sector models (see e.g.
Bolkesjø 2004) has provided possibilities to perform analyses related to, for example, marketing
(prices, costs), external conditions (subsidies, different rates of economic growth, currency
fluctuations), large-scale transport logistics, and location of the industry. Examples of such
analyses where biomass for energy purposes have been considered have been provided by e.g.
Bolkesjø et al. (2006) and Trømborg et al. (2007a, 2007b). The forest-sector approach, however, is
not considered further in this review.
3.1.2. Research challenges
Although the above-described Norwegian analyses in many respects capture a wide range of
challenges related to the estimation of available bioenergy resources, there are many aspects that
could be considered to improve the estimates and provide new information. In addition to
productive forest areas, also non-productive forest areas and ‘other forested areas’ should be
included. The same applies to more to marginal areas, such as densely populated areas, pasture,
power lines, roadsides, and border zones surrounding agriculture land.
The environmentally oriented restrictions (ranging from e.g. forest reserves where no harvest is
allowed to management activities adapted to the Living Forest standards (Living Forest 2010))
should preferably be applied as spatially explicit (i.e. located to individual sample plots) instead of
as a generally imposed percentage reduction of areas or biomass as in previous analyses (see
section 3.2).
Quantifications of the potential availability of tree biomass resources should also aim at presenting
results according to different strata, depending on the purpose of the analyses. Different
36
�
geographical units, land-use and land-type categories, forest conditions, growing site
characteristics, and biomass fractions are examples of strata that could be considered.
Previously, also only a limited range of alternatives regarding harvesting systems (harvest residues
according to conventional timber harvesting and to some extent early thinning) have been
considered. Also, additional harvesting systems, such as whole tree harvesting and stump
harvesting, should be included. In particular, productivity and the associated costs related to the
harvesting systems should be investigated, and corresponding models should be developed (see
section 3.3).
3.2. Environmental considerations
3.2.1. Background and previous research
The same environmental restrictions and considerations apply to conventional timber production
as to production of biomass for energy purposes. The environmental restrictions are materialized
in legislation and certification regimes focusing on forestry in general. Harvesting of forest
residues (branches and tops) and below-ground biomass (stumps and roots) for energy purposes,
however, raise some additional concerns (see e.g. Bohlin & Roos 2002, Rudolphi & Gustafsson
2005, Stupak et al. 2007, Lattimore et al. 2009, Lauren et al. 2008, Rabinowitsch-Jokinen &
Vanha-Majamaa 2010). Since a large proportion of all nutrients in trees is located in their needles,
removing residues will reduce nutrient supply to soils, and in the long run this may increase the
risk of nutrient imbalance and reduced productivity (Raulund-Rasmussen et al. 2008). Removing
residues may also affect biodiversity and change species composition (Åström et al. 2005). The
removal of stumps and roots raises concerns regarding risks of landslides, less organic matter in
soils, and decreases in substrates important for some species (Stupak et al. 2007, Lindhe 2009). A
few European countries have developed guidelines or recommendations for harvesting forest
residues and below-ground biomass (Stupak et al. 2007), but currently there are no guidelines in
Norway. However, in future it is expected that such guidelines will be developed as a part of
certification standards.
Environmental considerations in previous Norwegian assessments of potential biomass resources
for energy have in general been relatively simple, and only very rough estimates of the impacts
have been done (OED 1997, NOU 1998, KanEnergi 2003, 2007). Also, the assessments by
Langerud et al. (2007), Hobbelstad (2007a, 2007b), and Gjølsjø & Hobbelstad (2009) were
37
�
relatively simple, and few attempts were made to make the restrictions spatially explicit or to
distinguish between different types of restrictions.
Regarding assessments of available timber resources, comprehensive analyses of environmentally
oriented restrictions at national and regional levels as well as at property level have been done
(Hoen et al. 1998, Eid et al. 2001, 2002). This work was done as a part of the Living Forest project
(see section 3.2.2.), and since the standards of the certification regime (Living Forest 1998, 2001)
had not yet been agreed, assumptions regarding the details of the standards had to be made. The
analyses were done using Gaya-SGIS and included maintaining minimum proportions of old-
growth forest, prolonged rotation cycles, treatments in buffer zones related to key biotopes, water
bodies (rivers, lakes, etc.) and mires. The analyses at national and regional levels were based on
NFI sample plots. The assumptions made for the treatments related to the environmentally
oriented restrictions (Hoen et al. 1998), however, were mostly imposed according to a general
description of the forest contitions, and not based on spatially explicit prescriptions from
observations in the field. For the analyses made at the property level (Eid et al. 2001), the analyses
were based on observations of, for example, border zones and key habitats derived from maps or
observed in field. However, the assumptions were still based on the expected standards and not on
those finally agreed upon.
Later, Ask et al. (2005) and Bergseng et al. (2009) performed analyses using Gaya-SGIS
comprising data from Ski Municipality forest based on the assumptions and prescriptions of the
silvicultural treatments from current Living Forest standards (Living Forest 2010). To date,
however, the new standards have not been operationalized for NFI sample plots. The following
sections provide descriptions and discussions regarding the most important environmentally
oriented restrictions, and the impact these restrictions may have on the utilization of timber and
biomass resources in Norway.
3.2.2. Environmentally oriented instruments and their impact on forest production
Environmentally oriented instruments in Norway range from ‘hard’ instruments related to
legislation and regulations to certification regimes and standards, and to ‘soft’ instruments such as
preservation administered by individual forest owners. Important Acts influencing forest
production in Norway are the Act relating to the management of biological, geological and
landscape diversity (Nature Diversity Act, Act No. 100 of 19 June 2009), the Act relating to
forestry (Forestry Act, Act No. 31 of 27 May 2005), the Act relating to outdoor recreation
38
�
(Outdoor Recreation Act, Act No. 16 of 28 June 1957), the Act relating to planning and building
(The Planning and Building Act, Act No. 77 of 14 June 1985), and the Act relating to river
systems and groundwater (Water Resources Act, Act No. 82 of 24 November 2000).
The Living Forests standard has been implemented as basis for forest certification in Norway.
Living Forests is a collaborative project between various stakeholders (forest owners, forest
industry, trade unions, outdoor recreation organizations, and environmental organizations) to
develop a framework for sustainable forest management. The main aim of the project is to achieve
a balance between the three aspects forest production, environment, and social interests. The
outcome was the published standard ‘Living Forests: Standard for sustainable forest management
in Norway’. The current standard is documented by Living Forest (2010). Most of the Norwegian
forests are now certified according to the standard, either as individual properties or by group
certification. The criteria applied as Living Forest standards fulfil the requirements of certification
according to PEFC (Programme for the Endorsement of Forest Certification). A working group is
currently preparing a Norwegian Forest Stewardship Council (FSC) standard. Until this is ready it
is possible to certify forests through a ‘generic’ standard. However, the FSC is not widespread in
Norway to date.
Nature Diversity Act (see Table 3, first four columns)
The most important legislation restricting forest production in Norway is the Nature Diversity Act
(Act No. 100 of 19 June 2009). There are four relevant categories of site protection under this act:
Nature Reserve (§ 37), National Park (§ 35), Protected Landscape Area (§ 36), and Habitat
Management Area (§ 38). There is also a fifth category of protection, Nature types (§ 52), which
not yet has been put into practice. This category may influence forest production in future.
In general, protection is established by the King in Council by means of management regulations
specific for each site. Nature Reserves and National Parks provide a strong protection regime.
Normal timber production activities are not allowed, and these two categories are therefore
considered as fully protected. Protected Landscape Areas and Habitat Management Areas both
imply restrictions on the utilization of the forest resources. The regulations, however, vary
between sites and the prescriptions of treatments are partly formulated as ‘recommendations’ and
partly as ‘requirements’. The prescriptions may be connected to continuous cover forestry (i.e.
selective cutting required) and regeneration method and size of the clear-cut area (e.g. 0.2 ha). In
addition, road construction is normally prohibited. If clear cutting is not allowed for such areas,
39
�
the immediately available harvest quantities are obviously reduced. Applying selective cuttings,
however, does not necessarily reduce the production and harvests in the long run (Lexerød 2007,
Andreassen 1994). This depends on local forest conditions and suitability for applying selective
cutting (Lexerød & Eid 2006). The impact of the regulations also varies according to how strongly
they are formulated. If, for example, natural regeneration is a total requirement, the impact may be
high, while it may be lower if natural regeneration in the regulations is stated as ‘the preferred
method when suitable’. To what extent prohibition of road construction reduces harvests is an
economic issue that depends on terrain conditions and distance to presently existing roads.
Living Forest Standard and legislation (see Table 4, first four columns)
The Forestry Act (Act No. 31 of 27 May 2005) imposes several restrictions on forest management.
Most restrictions are materialized in the Living Forest Standard (Living Forest 2010). Also other
Acts are closely related to and partly work parallel with the Living Forest standard. The most
important requirements in the standard, namely buffer zones, protection of key habitats, retention
of living trees and dead wood, management adaptations for outdoor recreation activities,
management adaptations in mountainous areas, forest structure, game habitats, and cultural
monuments, are described in the following:
- The Water Resources Act (Act No. 82 of 24 November 2000) requires a natural belt of
vegetation to be maintained along the banks of river systems with perennial flow (§ 11). The
Living Forest Standard, section 12, states that buffer zones are required to be maintained or
developed not only along river systems, but also along smaller streams, bogs, lakes, and cultural
landscapes. The background for this standard is the high number of vital ecological functions
related to biodiversity, water quality, landscape, and outdoor recreation served by the buffer
zones.
The basic buffer zones’ width given in the Living Forest Standard is 10–15 m. However, the
width should be adapted to local conditions and may vary within individual buffer zones, i.e. it
may be narrowed down to 5 m or widened up to 30 m. Maintenance of existing zones or
development of new zones implies that harvesting is allowed to a certain extent, i.e. maintenance
by selective cutting in existing buffer zones or a more intensive harvesting in a one-storied forest
where suitable trees are left to develop a new buffer zone. There are a few exceptions where the
requirement of a buffer zone may be ignored, e.g. for the purposes of outdoor recreation (Living
Forest Standard) and constructions connected to river systems (Water Resources Act). Thus,
40
�
there are three aspects related to buffer zones that are challenging when estimating the impact on
forestry activities: 1) varying buffer zone widths, 2) different extents of harvesting allowed
within the buffer zones, and 3) exceptions where the requirements may be ignored.
- Key habitats are defined as especially valuable sites from a conservation point of view, normally
based on the presence of rare forest types or unusual structural elements, or the presence of rare
species (Sverdrup-Thygeson 2002). According to the Living Forest Standard, key habitats ‘shall
be inventoried, selected, documented and specified on maps’. There is no fixed general level for
how much of a productive forest area that should be specified as a key habitat. The inventory
instructions and the selection among the key habitats identified in the field vary according to
local conditions at municipality level and the abundance of certain habitats at national level. For
habitats that are rare at national level, but which frequently appear locally, only a small part may
be finally specified as a key habitat, whereas all of them may be finally specified if the frequency
is very low locally.
Two methods for identifying the key habitats in field are officially approved: Miljøregistrering i
Skog (MiS) (Hotspot Inventory in Forests) and the Siste Sjanse-metoden (SiS) (Last Chance
Method). The approaches of the two methods for identifying key habitats differ. While MiS
focuses on habitats, i.e. certain forest types and structural elements, SiS focuses on the presence
of indicator species. The Norwegian NFI has applied the MiS method for its sample plot
inventory.
In general, the impact of key habitats on immediate harvest as well as long-term production is
high. The key habitats may be totally preserved or some restriction may be imposed (i.e. some
harvesting may take place). If harvesting is allowed, this has to be done in a way that does not
weaken the conditions for biodiversity or that improves the conditions for biodiversity. In
practice, this implies selective cutting. Thus, there are at least three aspects related to key
habitats that are challenging when assessing the impact on forestry activities: 1) two different
approaches to fieldwork, 2) the selection process for final specification of key habitats, and 3)
varying levels of harvesting.
- According to the Living Forest Standard, section 8, an average of 10 living wind-resistant trees
ha-1 must be left after harvesting. These retention trees should normally be selected among the
oldest, and thus often the largest trees. The retention trees can also be part of buffer zones. Thus,
41
�
the actual effect of this requirement may be less than the effect of leaving 10 large trees ha-1 after
harvesting.
- Also dead wood, such as standing dead deciduous trees, large dead pines, natural high stumps,
and dead wood older than five years lying on the ground, should normally not be harvested
(Living Forest Standard, section 8). Some of this wood could potentially be used for pulpwood
or energy purposes. The general impact of this requirement, however, is low.
- In Norway the rights of every person to unimpeded access to the forests are strong. This is
regulated through the Outdoor Recreation Act (Act No. 16 of 28 June 1957), section 2, which
state that ‘any person is entitled to access to and passage through uncultivated land at all times
of year, provided that consideration and due care is shown’ (§ 2). The same applies to passage
on horseback or with a packhorse, sledge, and bicycle on roads and paths. The Act also includes
the right to pick berries and mushrooms. Some areas will naturally be more used for outdoor
recreation than others, e.g. some coastal and mountainous areas. In such cases, the recreation
areas can be secured by the State through compensation of the forest owner, or by means of a
municipal zoning plan (The Planning and Building Act, Act No. 77 of 14 June 1985). In addition
to forest areas set aside especially for outdoor recreation, general considerations should be taken
regarding harvesting activities to secure outdoor recreation (section 7 in the Living Forest
Standard).
All of the above-mentioned situations imply restrictions on forestry activities. The practical
impact will most likely differ between the type of restriction (forest areas set aside especially for
outdoor recreation as opposed to smaller adjustments of forest management for outdoor
recreation).
- Two kinds of restrictions exist related to forest areas adjacent to mountains: those stipulated
under the Forestry Act and those by the Living Forest Standard. The Forestry Act defines
‘protective forests’ as areas where the forest serves as protection zone for other forest or provides
protection against natural damage (§ 12). This applies to areas near mountains or on the coast,
where forests are vulnerable. According to the Forestry Act, the County Agricultural Committees
have the authority to issue regulations governing ‘protective forests’. Rules are normally
provided concerning the obligation to report to the authorities before any harvesting is done.
42
�
Also, the Living Forest Standard, section 6 – Mountain forest, regulates forest management in
mountain areas. The intention is to secure the biodiversity and recreational values of mountain
forests. According to section 6, the lower border of mountain forests is identical to the border of
protective forests in mountain areas, as defined in the Forestry Act. Thus, these areas will
overlap.
In general, harvesting in mountainous areas should be done in a way that promotes or maintains a
mature forest character. The harvesting method for spruce should as far as possible follow the
‘mountain forest selection system’, which requires that after harvesting the forest is still
classified as development class IV or V. Small clear-cutting areas regenerated with seed trees are
allowed in pine forests. There are at least two aspects related to mountainous forests that are
challenging when assessing the impact on forestry activities: 1) protective forest borders are not
available at a national scale, and 2) varying levels of removals in the harvesting.
- The Living Forest Standard, section 19 – Forest structure, requires that at any given time at least
30% of Norwegian forests should have a forest structure that is beneficial to species living in
mature forests. This requirement does not apply to the property level, and measures can only be
taken if the data from NFI show that the proportion of development class IV and V at county
level is less than 30%. The practical impacts on the activities are therefore rather low.
- Section 14 – Landscape plan, in the Living Forest standard is intended to secure that forest
management takes care of interests across stand and holding boundaries. Included here are
considerations regarding game habitats and mating areas for capercaillie (wood grouse, Tetrao
urogallus). The considerations are the same for all forestry activities, but there is no requirement
for developing a landscape plan for small properties (Sverdrup-Thygeson et al. 2002). However,
smaller forest properties known game habitats will, in most cases, be included in the forest
management plan. A specific requirement may, for example, be a 50 m buffer zone around nests
belonging to goshawks (Accipiter gentilis), where no harvesting is allowed.
- All monuments and sites originating from before 1537 and all Sami monuments and sites more
than 100 years old are automatically protected (Act Concerning the Cultural Heritage, Act No.50
of 9 June 1978). The Living Forest standard, section 13, concerns the protection of cultural
monuments and aims to secure all protected and other valuable monuments and sites in
Norwegian forests. Planting, soil scarification or other measures that might pose a threat to a
43
�
monument or site is prohibited. Identification of cultural monuments in field may be a challenge
because not all of them are surveyed and mapped. Cultural remains are not recorded by the
Norwegian NFI.
Other regulations (see Table 5, first four columns)
- The Ministry of Environment apply INON-areas as a tool for preserving existing wild areas in
Norway (Parliamentary bill 1 S (2009-2010)). INON-areas are described as areas without major
infrastructure development and defined as being 1 km or longer in distance from technical
installations such as roads, railways and power lines. The general aim is to protect such areas
from new technical installations. Regular harvesting is not regarded as a ‘new technical
installation’, but forest roads are. The forest management of such areas will therefore be
indirectly influenced because road construction is not allowed. The impact of these regulations
will vary according to whether it is profitable or not to carry out regular harvesting activities
within the present road system. The profitability depends on local conditions, such as distance
from harvesting site to nearest road, and terrain factors such as steepness, soil conditions, cliffs,
etc. (see also Eriksen et al. 2004).
- In forest areas close to densely populated areas special restrictions may be imposed. Such
restrictions may vary from being legislative to being administrative. The area surrounding Oslo,
for example, is regulated under the Act relating to nature areas in Oslo and nearby municipalities
(Act No. 35 of 5 June 2009). Also, in other forested areas owned by the municipalities,
management is often adjusted to meet the demands of the public (e.g. Ski Municipality forests,
Trondheim Municipality forests). These adjustments may be anchored both administratively and
politically, and thus strongly binding. The impact of these restrictions may vary considerably.
The practical silvicultural prescriptions, however, are usually connected to limitations on clear-
cutting or recommendations for selective cutting.
- Also other restrictions and regulations not previously described may have an impact on forestry
activities. Some examples are restrictions related to forest areas that recently have been burned,
general adjustments of forest management for wildlife and landscape reasons, administratively
protected private areas, and restrictions related to some nature types based on registrations at
municipal level.
44
�
According to the Living Forest Standard, section 4, ‘at least 5% of productive forest areas shall be
managed as areas of ecological importance’. Such areas include key habitats (provided by Living
Forests Standard) and nature reserves and national parks (provide for under the Nature Diversity
Act). If these elements do not fulfil the 5% requirement, a forest owner may make up the deficit
by selecting areas from different forest types (e.g. calcareous forests, swamp forests, broadleaved
temperate forests, pasture woodland, and other types). Also buffer zones in development classes
IV and V may be included. Non-productive forest land may constitute up to 25% of the 5% area
requirement. Thus, requirement stipulated by section 4 of the Living Forest Standard will most
likely not influence the forest activities directly.
3.2.3. Challenges in operationalizing impacts
In addition to the actual management prescriptions related to the environmentally oriented
restrictions and the impact they may have on immediate harvests and long-term production, many
challenges need be faced to ensure that the prescriptions are operational when assessing available
biomass resources. In the following, some challenges related to large-scale assessments based on
Norwegian NFI sample plot data are discussed (see Tables 3, 4, and 5, last two columns).
The basic requirement for operationalizing is available spatially explicit information on the type of
restriction. For some of the restrictions this is straightforward because the information is directly
recorded by the NFI in the field or recorded by means of digital map data. This applies to
restrictions related to the Nature Diversity Act and for most restrictions related to the Living
Forest Standard. However, an exception regarding the Living Forest Standard is ‘Mountain
Forests’, for which national digital maps do not exist. Similar problems apply also to forests near
densely populated areas, game habitats, cultural remains, and a number of other restrictions. In
general, where spatial information is not available, restrictions need to be handled in a ‘pool’
where the impacts are imposed subjectively and are spatially inexplicit, subsequent to all other
restrictions which have been implemented.
Depending on whether the aim is to assess the present total biomass resources only, without
including temporal aspects, or whether the aim also is to include temporal aspects by means of
long-term analyses applying a decision support system, the approach to assessments will differ. In
the former case, the impact on the immediate harvest and long-term production has to be assessed
jointly by means of subjective judgements. For restrictions that impose a full protection regime
(e.g. Nature Reserves or National Parks) this is straightforward because the present biomass will
45
�
be reduced by 100%. For all other restrictions, where some activities actually are allowed, the
reduction should vary according to the type of activities allowed and the joint impact on
immediate harvests and long-term production.
For long-term assessments it is necessary to apply a decision support system comprising a forest
simulator describing growth and different silvicultural treatments over time. Gaya-SGIS has
previously been used as tool for such assessments and most restrictions described in Tables 3, 4
and 5 (last column) can be applied as long as spatial information at plot level is available (Hoen et
al. 1998, Ask et al. 2005). Since Gaya-SGIS is non-spatial in the sense that it does not allow
adjacency restrictions, e.g. for reducing forest fragmentation using the core area concept (e.g.
Øhman & Eriksson 1998), limitations on the clear-cutting area cannot be appropriately handled.
Most remaining restrictions can be handled appropriately with the present version of Gaya-SGIS,
although some ‘fine tuning’ according to the present regulations and standards may be necessary.
46
�Tabl
e 3.
Env
ironm
enta
lly o
rient
ed re
stric
tions
acc
ordi
ng to
the
Nat
ure
Div
ersi
ty A
ct.
Form
al b
ackg
roun
d Fu
nctio
n D
escr
iptio
n Im
pact
Sp
atia
lly e
xplic
it
info
rmat
ion
at
NFI
plo
t lev
el
App
licab
le a
t
plot
leve
l for
Gay
a-SG
IS
Im
med
iate
harv
est
Long
-term
prod
uctio
n
Nat
ure
Div
ersi
ty A
ct
Nat
ure
Res
erve
V
ery
few
or n
o ac
tiviti
es a
llow
ed
Full
Full
Yes
Y
es
Nat
ure
Div
ersi
ty A
ct
Nat
iona
l Par
k V
ery
few
or n
o ac
tiviti
es a
llow
ed
Full
Full
Yes
Y
es
Nat
ure
Div
ersi
ty A
ct
Prot
ecte
d
Land
scap
es A
rea
No
road
con
stru
ctio
n
Lim
itatio
n on
cle
ar-c
uttin
g ar
ea
Req
uire
d or
reco
mm
ende
d se
lect
ive
cutti
ng
Req
uire
d or
reco
mm
ende
d na
tura
l reg
ener
atio
n
Som
e
Hig
h
Hig
h
Som
e
Som
e
Som
e
Som
e
Som
e
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Nat
ure
Div
ersi
ty A
ct
Hab
itat
Man
agem
ent A
rea
No
road
con
stru
ctio
n
Lim
itatio
n on
cle
ar-c
uttin
g ar
ea
Req
uire
d or
reco
mm
ende
d se
lect
ive
cutti
ng
Req
uire
d or
reco
mm
ende
d na
tura
l reg
ener
atio
n
Som
e
Hig
h
Hig
h
Som
e
Som
e
Som
e
Som
e
Som
e
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
47
�Tabl
e 4.
Env
ironm
enta
lly o
rient
ed re
stric
tions
acc
ordi
ng to
Liv
ing
Fore
st st
anda
rds a
nd le
gisl
atio
n.
Form
al b
ackg
roun
d Fu
nctio
n D
escr
iptio
n
Impa
ct
Spat
ially
exp
licit
info
rmat
ion
at N
FI
plot
leve
l
App
licab
le a
t
plot
leve
l for
Gay
a-SG
IS
Im
med
iate
harv
est
Long
-term
prod
uctio
n
Livi
ng F
ores
t Sta
ndar
d
Wat
er R
esou
rces
Act
Bor
der
zone
s
No
clea
r-cu
tting
, sel
ectiv
e cu
tting
allo
wed
H
igh
Som
e Y
es
Yes
Livi
ng F
ores
t Sta
ndar
d K
ey h
abita
ts
No
harv
est,
sele
ctiv
e cu
tting
som
etim
es a
llow
ed
Hig
h H
igh
Yes
Y
es
Livi
ng F
ores
t Sta
ndar
d
Ret
entio
n of
livin
g tre
es
At l
east
10
trees
ha-1
left
afte
r har
vest
ing
Som
e So
me
Yes
Y
es
Livi
ng F
ores
t Sta
ndar
d D
ead
woo
d C
erta
in q
ualit
ies l
eft a
fter h
arve
stin
g
Low
Lo
w
Yes
Y
es
Livi
ng F
ores
t Sta
ndar
d
Out
door
Rec
reat
ion
Act
Plan
ning
and
Bui
ldin
g A
ct
Out
door
recr
eatio
n
Gen
eral
con
side
ratio
ns fo
r all
activ
ities
So
me
Som
e Y
es (M
aps r
equi
red)
Po
ssib
ly
Livi
ng F
ores
t Sta
ndar
d
Fore
stry
Act
Mou
ntai
n
fore
st
Mou
ntai
n fo
rest
sele
ctio
n of
spru
ce
Lim
itatio
n on
cle
ar-c
uttin
g ar
ea p
ine
Hig
h
Hig
h
Som
e
Som
e
Yes
(Map
s req
uire
d)
Yes
(Map
s req
uire
d)
Yes
No
Livi
ng F
ores
t Sta
ndar
d
Fore
st
stru
ctur
e
Min
imum
30%
at a
ny ti
me
in d
evel
opm
ent
clas
ses I
V–V
at c
ount
y le
vel
Low
Lo
w
Yes
Y
es
Livi
ng F
ores
t Sta
ndar
d
Gam
e
habi
tats
No
harv
est
Hig
h H
igh
Yes
(Map
s req
uire
d)
Yes
Livi
ng F
ores
t Sta
ndar
d
Cul
tura
l Her
itage
Act
Cul
tura
l
mon
umen
ts
and
site
s
Har
vest
or o
ther
mea
sure
s dis
figur
ing
any
mon
umen
t or s
ite is
pro
hibi
ted
Low
Lo
w
Map
s req
uire
d Y
es
Tabl
e 5.
Env
ironm
enta
lly o
rient
ed re
stric
tions
acc
ordi
ng to
oth
er re
gula
tions
.
48
�Form
al b
ackg
roun
d Fu
nctio
n D
escr
iptio
n
Impa
ct
Spat
ially
exp
licit
info
rmat
ion
at
NFI
plo
t lev
el
App
licab
le a
t
plot
leve
l for
Gay
a-SG
IS
Im
med
iate
harv
est
Long
-term
prod
uctio
n
Wild
are
as (I
NO
N)
Spec
ific
cons
ider
atio
n
No
road
con
stru
ctio
n So
me
Som
e Y
es
Yes
Act
rela
ting
to n
atur
e ar
eas i
n O
slo
and
near
by m
unic
ipal
ities
Dec
isio
n by
mun
icip
ality
aut
horit
ies
Spec
ific
cons
ider
atio
ns
Rec
omm
ende
d se
lect
ive
cutti
ng
Lim
itatio
n on
cle
ar-c
uttin
g ar
ea
Som
e
Som
e
Som
e
Som
e
Yes
Yes
Yes
No
Adm
inis
trativ
e pr
otec
ted
area
s Pr
otec
tion
of
fore
st a
reas
Var
ying
. Nor
mal
ly n
o
harv
estin
g.
Hig
h H
igh
Map
s req
uire
d Po
ssib
ly
49
�
3.3. Economic considerations
3.3.1. Background
The supply chain of biomass for energy purposes comprises a number of operations starting in the
forest stand and ending at the power plant. The qualities and quantities required by the end-users
to a large extent determine what kind of operations that are needed to be done. There are a lot of
challenges related to logistics, technical solutions and production regarding storing and woodchip
cutting at roadside and transport from roadside to end-user (see e.g. Junginger et al. 2005, 2006,
Laitila 2006, Gjølsjø & Belbo 2008). This part of the biomass supply chain will not be considered
in the present review.
Assessments of available timber quantities in Norway have been based on traditional
considerations regaring gross timber value and harvesting costs. The timber quantities for a certain
area have been regarded as non-available (‘0-areas’) if the difference between gross timber value
and harvesting cost is negative, while they have been regarded as available if the difference is
positive. Previous assessments of large areas have been based on this concept (see e.g. Hoen et al.
1998, 2001, NIJOS & NORSKOG 1999). Although there are many questions and uncertain factors
related to such assessments (Bollandsås et al. 2004a, 2004b), estimations are quite straightforward
regarding timber production due to the availability of models reflecting timber values in the
market (Blingsmo & Veidahl 1992, Lexerød & Gobakken 2002) and models describing
productivity and costs according to different forest conditions and harvesting methods (e.g. Dale
et al. 1993, Dale & Stamm 1994, Eid 1998). However, many of these models are rather old and
some updating is needed.
A similar approach, based on biomass values in the market and models for productivity regarding
harvesting methods, could also be applied when considering biomass production. The biomass
market for energy in Norway is rather ‘immature’, and a structured and well-documented price
system similar to what we see for timber (saw timber, pulpwood) does not exist. One may, for
example, expect different prices according to different fractions (harvesting residues, i.e. branches
and tops, whole tree, stumps), tree species, and possibly treatment (e.g. storing in forests for
drying purposes). Until a price system is established in Norway the only option is probably to
assume a certain standardized price per ton at roadside independent of fraction, tree species or
treatment. Harvesting methods related to bioenergy production, i.e. the operations in the supply
chain comprising harvesting and transport of biomass to roadside, are discussed in the following.
50
�
3.3.2. Biomass harvesting methods
The supply of most biomass for energy production in Norway today and in the near future relies
heavily on forest residues (branches and tops) from traditional timber production based on clear-
cutting. Theoretically, however, there are a number of methods that could be applied to harvest the
biomass, either where timber production is the main aim of the operations and biomass quantities
is a result of this, or where biomass is the main production aim and all quantities are used for
energy purposes. Table 6 gives an overview of these possibilities according to different
silvicultural treatments.
Table 6. Timber and biomass production possibilities according to silvicultural treatments. Silvicultural treatment Aim of production Timber fraction Biomass fraction
Clear-cutting Timber
Energy
Stem
-
Harvest residues, stumps
Whole tree, stumps
Seed tree establishment
and cutting
Timber
Energy
Stem
-
Harvest residues
Whole tree
Shelterwood establishment
and cutting
Timber
Energy
Stem
-
Harvest residues
Whole tree
Thinning Timber
Energy
Stem
-
Harvest residues
Whole tree
Selective
cutting
Timber
Energy
Stem
-
Harvest residues
Whole tree
Young growth tending Timber
Energy
None
-
-
Whole tree
For clear-cutting, where the main aim is timber production, biomass for energy may come from
harvest residues and stumps, while it may come from whole tree and stumps if the main
production aim is energy. The same options exist, with the exception of stump harvesting, for seed
tree and shelterwood establishment and cutting, conventional thinning, and selective cutting.
Harvesting stumps for these silvicultural treatments would not be possible due to existing trees or
newly recruited seedlings. For all the silvicultural treatments, except young growth tending, there
is, at least theoretically, a competition between producing timber and biomass for energy. For
young growth tending, this competition does not exist: either the felled trees remain in the forest
or they may be used for bioenergy purposes.
Since there is competition between timber production and biomass production for energy, all
silvicultural options and the corresponding harvesting and forwarding methods should be
51
�
compared with respect to productivity and profitability (to compare profitability, also the prices of
biomass and timber need to be considered). Models describing production for harvesting and
forwarding according to different forest conditions (e.g. volume ha-1 harvested, size of trees
harvested, number of trees ha-1 prior to harvest, distance to roadside) and harvesting methods exist
for all silvicultural treatments regarding timber production as described in Table 6 (Dale et al.
1993, Dale & Stamm 1994, Eid 1998). Currently, no corresponding models for harvesting of
biomass exist in Norway. In the following, research on production related to the most important
biomass harvesting options, mainly from Sweden and Finland, are discussed.
- Clear cutting, seed tree and shelterwood establishment and cutting
Few studies exist for estimating productivity or the costs of collecting harvest residues after
conventional clear-cutting for timber production. The most relevant study for Norwegian
conditions is probably of the harvesting of loose residues from Norway spruce dominated stands
in Finland (Nurmi 2007). Productivity studies were carried out to study the effect of a single-grip
harvester work method on the forwarder productivity when collecting the residues. The study
comprised 65 effective machine hours (E0) and the operations were divided into loading,
forwarding, unloading, and driving unloaded. The results were presented partly as average values
for production according to different sub-operations and partly as functions describing how the
production varied according to different conditions. Rørstad et al. (2010) applied these results and
developed a set of models where time consumption (h/ton) was estimated for the following
operations:
- transport (average for forwarding and driving back unloaded) based on a function with hauling
distance (km), driving speed (km/h) and load size (ton) as independent variables
- loading based on a given average productivity (ton/ h) calibrated according to residue density
(ton/ha)
- unloading according to average productivity (ton/h) (Nurmi 2007)
The cost (NOK/ton) was finally calculated by multiplying time consumption (h/ton) by machine
cost (NOK/h).
The bundling system, as an alternative to the above-described harvest method for loose residues,
has been studied by Kärhä & Vartiamäki (2006). In the bundling system, the logging residues left
by a harvester are collected and fed into a bundler to produce compact bundles which are then
52
�
forwarded to roadside. The study revealed that the cost of the bundling chain was higher than the
cost of the loose residue chain when reviewing the total costs from harvest to roadside.
A few relevant studies on stump harvesting also exist. Karlsson (2007) in Sweden presented
interesting results for lifting as well as forwarding of stumps based on time consumption studies.
Average values and models for predicting production and cost were presented. The study was
based on one spruce-dominated stand. Stump harvesting has also been studied by Lazdi�š et al.
(2009) in Latvia, and results for production and costs based on three plots presented. Laitila
(2006) have developed a Microsoft Excel program for calculating the production and cost of
different harvesting methods. Stump harvesting is included in this program, where basic
productivity calculations are based on a study by Laitila & Asikainen (2004).
Although clear-cutting in old mature stands, where the main aim is energy production (Table 6),
i.e. harvesting of whole trees, probably is of marginal practical importance for most forested areas,
the method might be of interest for forested areas such as old pasture and border zones
surrounding agricultural land. No productivity studies have been found regarding old mature
forests. For smaller trees, however, several studies exist (see below).
For seed tree establishment and cutting (Table 6) similar consideration as those for clear cutting
apply. Final cutting of the seed trees, where biomass from residues or whole trees is collected for
energy purposes, however, is probably not relevant because the biomass density ha-1 will be too
low. Low biomass density is probably also a problem when considering energy production for the
establishment as well as final cutting of shelterwood. No productivity studies have been found
regarding seed trees or shelterwood.
- Thinning and selective cutting
A lot of research has been done previously in Finland on production and costs related to thinning.
However, no research on the collection of loose forest residues from conventional thinning has
been found. Instead, the research has focused on different methods related to whole tree harvesting
(see e.g. Kärhä et al. 2005, Kärhä 2006, Ovaskainen et al. 2008) and whole tree bundle harvesting
(Jylhä & Laitila 2007, Laitila et al. 2009) regarding thinning in young stands. Kärhä (2006), for
example, studied two different harvesting systems, one with a separate harvester and forwarder
and the other with a harwarder, i.e. one machine combining harvesting and forwarding. Results
53
�
were presented on productivity and cost variation according to, for example, distance to roadside
and average tree size. No productivity models were presented.
Possibly the most interesting study regarding models for estimating productivity has been
provided by Laitila et al. (2009). Models were developed for all sub-operations related to the
transport of whole tree bundles to roadside. Laitila et al. also found that the productivity of the
bundle system was almost twice as high compared to conventional individual whole tree
forwarding.
- Young growth tending (and early thinning)
Laitila et al. (2007) studied the forwarding of whole trees after manual and mechanized felling in
pre-commercial thinning (the majority of the trees were smaller than 12 cm at breast height).
Forwarding after mechanized felling gave the highest productivity and lowest costs of the two
methods. Productivity models with grapple log size and density (m3 per 100 metres strip road) as
independent variables were developed. No results for the productivity and costs of the felling
operation were presented.
Laitila (2008) compared three systems for early thinning (i.e. the average size of trees was 0.030
m3): manual cutting and forwarding of whole trees, mechanized cutting and forwarding of whole
trees, and a system using a combined harvester and forwarder. The results for productivity and
costs according to different sub-operations were presented.
A study of biomass harvesting of small trees, where a certain amount of trees were left for further
growth, was provided by Björheden et al. (2003). A total of 10 different stands were included in
the study and the average diameter in the removals was between 3 cm and 10 cm. Productivity and
the costs of a number of different harvesting methods, from motor manual felling to fully
mechanized methods, were presented. The study focused on the removal of small trees in urban
areas, but may also be relevant for harvesting of biomass from ‘energy forest production’, i.e.
pioneer tree species in short-rotation intensive silviculture regimes (see also section 4.2).
3.3.3. Research challenges
As previously stated, there are many challenges related to logistics, technical solutions and
production regarding storing at roadside and transport from roadside to end-user. Quite often the
harvesting methods bringing biomass to roadside are also dependent on the technical solutions
54
�
selected for the remaining chain. It is therefore important to develop ‘a map’ describing the supply
chain of biomass from the forest stand to the power plant covering all options for operations and
technical solutions. Such a map should form the basis for future research on the harvesting and
logistical issues of the supply chain for biomass.
Ideally, productivity models covering all possibilities according to the different silvicultural
treatments listed in Table 6 should be developed. However, it is particularly important to study the
possibilities that combine the recovery of biomass residues from conventional clear-cutting for
timber production. Today, the biomass quantities harvested in this way are low. To be able to
reach the ambitious targets for future bioenergy production in Norway, it is quite obvious that
most quantities will need to be based on such residues. However, also whole tree harvesting, from
conventional thinning and from clear-cutting in areas not normally considered for timber
production (pasture, power lines, roadsides, border zones around agricultural land), and possibly
stump harvesting after conventional clear-cutting, needs to be considered in order to reach the
targets. For all these harvesting systems it is therefore necessary to develop productivity models
that capture variations related to terrain and location (e.g. slope and distance to roadside) and
variation related to forests (density measures such as volume, basal area or number of trees ha-1,
and dimensions, i.e. diameter, height and volume of individual trees).
There are currently few models developed for Norwegian conditions that can be applied to
estimate the productivity and costs of the harvesting systems. However, as demonstrated by
Rørstad et al. (2010), it is possible to approach this based on productivity data from the literature.
The large number of previous studies in Sweden and Finland constitute an excellent basis for
further development along these lines. However, it is also important to conduct research in
Norway regarding the time consumption and productivity of the harvesting systems, especially in
order to capture the terrain conditions (slope) present in Norway, but not in Sweden and Finland.
It also important to note that only a few production studies in Sweden and Finland focus on the
harvesting of residues after conventional clear-cutting. The only research found relevant for
Norwegian conditions is that provided by Nurmi (2007).
3.4. Decision support systems
Norway has long traditions in development and use of forest management decision support
systems. Such systems have regularly been applied for the purposes of research, teaching,
practical management planning for private and public properties, and by public authorities in
55
�
consequence analyses as a foundation for policy making. An extensive review of Norwegian
decision support systems and challenges related to applications and further development has
previously been done by Eid (2007). The existing decision support systems and their present and
future potential for applications related to assessments of biomass for energy purposes in Norway
are discussed in the following.
3.4.1. Existing decision support systems
Gaya-SGIS, AVVIRK-2000, and T (ree) are three software systems that over the past few years
have been frequently used and/or developed. The systems apply different methodological
approaches and serve different purposes for decision support. All three systems also currently
have functionality that may be useful for decision support directed towards biomass production for
energy purposes.
- Gaya-SGIS is a dynamic forest management system comprising a forest simulator and a linear
programming module integrated into a GIS environment. The system combines the forest
simulator, Gaya (Hoen & Eid 1990, Hoen 1996, Hoen & Gobakken 1997, Gobakken 2003),
which generate a range of realistic treatment schedules for each stand in the forest and their
corresponding net present values (NPV), and the linear programming (LP) model J (Lappi 2005),
which optimizes forest management given a set goal and restrictions. Gaya and J are integrated
in the ArcView GIS system. The system is non-spatial in the sense that it does not allow
adjacency restrictions. However, the GIS linking of Gaya-SGIS (Næsset 1997, Gobakken 2003)
enables the use of GIS to modify the data set for the simulator or to set restrictions for the
input/output matrix that transfers data from the forest simulator to the LP solver.
The forest simulator is developed from area-based empirical-statistical biological sub-models for
growth (Tveite 1977, 1978, Blingsmo 1984), mortality (Eid & Øyen 2003), and regeneration and
recruitment describing even-aged treatment alternatives (clear-cutting, seed tree and shelter
wood cutting). Some rough adjustments of the area-based biological models have also been
applied to facilitate selective cutting alternatives. The simulator describes the development over
time for all variables related to traditional timber production (e.g. basal area mean diameter,
mean height by basal, number of trees ha-1), and the corresponding timber assortments, timber
values (Blingsmo & Veidahl 1992, Lexerød & Gobakken 2002), harvesting costs, and net cash
flow. The simulator also estimates total biomass and biomass fractions applying equations
developed by Marklund (1987, 1988) and Petersson & Ståhl (2006). Since the simulator is area-
56
�
based, no information is available at individual tree level, and biomass is therefore estimated by
means of basal area mean diameter and number of trees ha-1 (see also section 2.1.2.).
Gaya was at an early stage developed to study carbon flows (Hoen & Solberg 1994, 1999), and
has since been substantially revised and currently comprises: (1) fixation of carbon in living
trees, (2) release of carbon from dead trees, litter, harvest residues and soil, (3) release of carbon
from wood products, and (4) saved greenhouse gas emissions (CO2, CH4, and N2O) from the use
of wood products instead of more energy intensive materials or fossil fuels (Raymer et al. 2009).
New elements also include annual production and mortality of needles, leaves, branches, and
fine roots and a process-based soil model. The latest improvement was the addition of cost
functions describing cost variations according to forest conditions and geographic dispersion
when harvesting residues from conventional timber harvesting (Rørstad et al. 2010).
- Avvirk-2000 is a forest management system comprising a forest simulator and a module
scheduling potential harvests paths at forest level (Eid & Hobbelstad 1999, 2000, 2005). There is no
optimization built in. The system should be operated heuristically, i.e. the user should compute
different alternatives through ‘intelligent’ manipulation of the assumptions and search for a few, but
satisfactory, solutions. No GIS functionality is currently linked to the system.
The forest simulator in Avvirk-2000 is similar to the one developed for Gaya, i.e. area-based
empirical-statistical biological sub-models for growth, mortality and regeneration/recruitment
describing even-aged treatment alternatives are applied. Also in this system adjustments of the
area-based models have been done to facilitate selective cutting (Eid & Hobbelstad 2005). The
basic biological performance of the system has been tested empirically based on historical
inventory and harvest-level records (Eid 2004, Eid & Hobbelstad 2006). This simulator also
provides quantities, timber values, harvesting costs, and NPV for traditional timber production.
Total biomass and biomass fractions are estimated in the same way as described for Gaya.
However, a comprehensive module for estimating harvest production and costs, including felling
and forwarding related to clear-cutting, seed tree cutting, thinning, and selective cutting has been
developed based on production studies and subsequent evaluations (Dale et al. 1993, Dale &
Stamm 1994, Eid 1998). No production models for estimating harvesting costs related to bio-
mass production are incorporated.
57
�
- T(ree) is a newly developed forest simulator based on sub-models for individual trees (Gobakken
2007, Gobakken et al. 2008). Currently, the simulator can only be applied for stand-level
analyses, but may in the future form part of a decision support system that can be applied for
large forest areas. The biological sub-models, i.e. distance-independent individual tree growth
(Bollandsås 2007) and mortality models (Eid & Tuhus 2001), and area-based regeneration and
recruitment models (Lexerød 2005, Lexerød & Eid 2005), are based on permanent sample plots
from the Norwegian NFI. The economical sub-models estimate timber values (Blingsmo &
Veidahl 1992, Lexerød & Gobakken 2002) and harvesting costs (Dale et al. 1993, Dale &
Stamm 1994) based on individual trees. For each management unit (i.e. stand), the simulator
produces treatment schedules with all feasible combinations of user-defined treatment and
regeneration options (pre-commercial thinning, thinning, different kinds of regeneration cutting
with different kinds of regeneration options, and selective cutting). A NPV is calculated for all
treatment schedules.
No functionality related to carbon flows or biomass estimation has been implemented to date.
The timber value estimation module in T has recently been rebuilt (see Lexerød et al. 2009).
Instead of, as previously, using functions to estimate timber value, T now applies species-
specific tree value matrices directly to estimate the timber value of harvested trees. The
generation of tree value matrices is based on log values for different combinations of log
diameter and length specified by the saw log industry and current pulpwood prices.
3.4.2. Research and development challenges
Gaya-SGIS and Avvirk-2000 are well suited for large-scale scenarios, for example related to
available biomass resources and carbon sequestration. The main strength of these tools is to
provide robust and appropriate results for larger areas (property level and higher). Gaya-SGIS and
Avvirk-2000 are less suitable for detailed analyses and descriptions of a wide range of silvicultural
options, especially those related to thinning and selective cutting. For this purpose T(ree),
comprising sub-models for individual trees, is potentially more suitable, for example regarding
density management, fertilization and energy forest issues (see sections 4.1 and 4.2). However,
more testing and evaluation of the basic biological models, and probably also development of new
models, are necessary before the program can be applied in analyses of biomass production
aiming especially for energy production.
58
�
Gaya-SGIS will probably be the main tool for large-scale and long-term assessments of biomass
resources in Norway in the future. The ultimate goal should be to develop Gaya-SGIS into a tool
that comprises a fully integrated approach to harvesting systems and associated costs as well as
prices for different timber assortments and biomass fractions, considering both traditional timber
production and production of biomass for energy purposes. This requires work regarding testing of
present functionality, the implementation of existing methods and models, the development of
new methods and models, and analyses with different purposes and at different scales. A few
challenges related to this are briefly described in the following.
- the present biomass estimation module is based on stand-level variables, i.e. basal area mean
diameter and mean height by basal area, while the applied biomass equations require tree-level
information. This procedure may cause biased estimates. Systematic testing of this problem to
quantify and evaluate the impact should be done. Alternatively, stand-level biomass expansion
factors could be developed (see section 2.1.2) and implemented in Gaya-SGIS.
- the present biomass estimation module is developed for analyses regarding carbon sequestration.
In carbon estimations the aim is to determine the total biomass while quantification of different
biomass fractions is less important. In estimations related to bioenergy it is important to quantify
the biomass that actually can be used for energy purposes. It is also important to distinguish
between different fractions since they have different potential in the production of energy.
- Gaya-SGIS quantifies biomass production in terms of kilograms of dry weight. The biomass,
however, also needs to be converted into heating value (kWh/ton). The heating value of biomass
depends on the composition of the biomass which may be different for different trees species and
different fractions of the trees. Usually, a fixed average heating value is applied. The heating
value also depends on the moisture content of the biomass, and the effective heating value varies
according to this content, which in turn varies according to the harvesting systems and
requirements of the end-users. A systematic review of these questions should therefore be carried
out in order to standardize the assumptions which need to be made for such analyses.
- the present forest simulator does not include cost productivity and calculations according to
mechanized timber harvesting methods (Dale et al. 1993, Dale & Stamm 1994, Eid 1998). A
module for such calculations should be developed. In addition, the work should start on in-
corporating productivity and cost models for bioenergy harvesting systems. As a first step,
59
�
Rørstad’s et al. (2010) ‘engineering’ approach to harvesting of residues, i.e. developing a set of
equations quantifying harvesting costs based on available productivity data, should be assessed
and possibly revised. A similar concept could also be applied to, for example, whole tree
harvesting and stump harvesting. As soon as Norwegian studies on productivity become
available, the results should be implemented in Gaya-SGIS by means of appropriate equations.
- the present forest simulator is able to quantify the gross value of the timber in a forest stand
based on models where the value varies according to the species and size of trees and the
prevailing price system. A similar price system for biomass does not exist in Norway, and
currently certain standardized prices per ton at roadside need to be implemented. However, as
soon as the market for biomass becomes more ‘mature’, i.e. as soon as price differentiation
appears according to, for example, different tree fraction, tree species or treatments, this should
somehow be included in the forest simulator.
- the forest simulator comprises sub-models for recruitment and regeneration, diameter growth and
height development, and mortality. Such models can always be calibrated or developed further
depending on new data that becomes available. The models describing regeneration, recruitment
and young growth phases are currently probably those with the highest potential for
improvements. An improved module describing the impact of fertilization on growth should
probably also be considered since such a treatment may be important in order to compensate for
nutrient losses associated with biomass harvesting.
60
�
4. Biomass production 4.1. Silvicultural methods in forestry
Traditional forest management is currently optimizing other values than the volume of biomass
production. Under current economic conditions in Norway it is most profitable to grow individual
trees to a minimum dimension for industrial use. The preference given to individual tree
dimension is always at the expense of reduced total stand volume and biomass production (Long
et al. 2004). Traditional forest management and biomass production are therefore in conflict. In
addition, the paper industry and bioenergy supply chains are competing over the same resources.
Even though in conflict, the production of wood for energy and industrial use will also in the
future be combined in the same stands and it is therefore important to design silvicultural systems
that optimize all goals simultaneously (Heikkilä et al. 2009). The main objective of silvicultural
methods adapted to management objectives that include bioenergy harvesting is to increase the
total biomass production (Egnell 2009). Many of the silvicultural methods used in forestry to
maximize the yield of biomass for fibre also apply if the same biomass or other biomass
components are used for energy rather than fibre (Ståhl 2009). Harvesting other biomass
components, such as branches, tops and stumps that have not been used previously, is another
option. It is not rationale to increase the proportion of such biomass components at the expense of
the more valuable stem biomass. Silvicultural methods have to increase the production of all
biomass components simultaneously in order to maximize the total biomass production.
A number of silvicultural methods to increase biomass production in the context of combined
production of wood for energy and industrial use are already well known because they apply in a
similar way matter if volume production is maximized for industrial wood production only. Few
modifications are necessary in order to adapt them to the combined production of the future. Due
to the fact that the maximization of total volume production has been neglected at the expense of
value production during recent decades, many of the research results are either outdated today or
qualitative rather than quantitative, or are based on experiences outside Norway. The silvicultural
methods that are known to significantly increase biomass production are density management, use
of fast growing pioneer species, and fertilization. We will therefore concentrate our review on
these three methods. Other methods discussed in less detail here are species selection (differences
in productivity on different sites, adaption to environmental changes, exotic species), genetic
improvement, intensive stand establishment (soil preparation, vegetation control, use of natural
61
�
regeneration in combination with planted seedlings, shelter, pine weevil control, reduced elk
browsing, use of stump sprouts), avoidance of root rot damage, and ditching (Ståhl 2009).
All silvicultural methods that are used to intensify management in order to increase biomass
production need to be carefully evaluated for their economic consequences. Due to the low prize
for bioenergy not all investments in intensified silviculture are profitable. Some methods might
give benefits in terms of increased biomass production for bioenergy and increased value of the
wood produced for industrial use. In order to evaluate the consequences of different management
options better growth models which predict not only the stem volume but also the biomass of all
tree components and different wood properties are needed. This is currently the largest research
challenge and needs long-term effort to be realized.
4.1.1. Density management
Density management is a simple method to increase biomass production because density has a
large effect on total wood production per area. Management affects density during establishment,
either by planting at a certain density or by preparing for natural regeneration, or during thinning
operations. For either of these methods, no quantitative decision support tools are available in
Norway that describe the effect of density on growth and that could help managers to decide on
optimal density. Practical density management during thinning operations is today mostly
accomplished by harvester drivers, who have little guidance on density targets and actual density.
The consequences of different thinning intensities on growth and yield are best projected using
individual-tree based growth models that are calibrated based on data representing a range of
different thinning intensities. Models of this type have been developed in Norway (Bollandsås et
al. 2008) but are not operational yet. Stand-level growth models for Norway have been shown to
underpredict growth after thinnings (Braastad & Tveite 2000). Thinning guidelines currently used
in Norway (Myklestad et al. 2006) are therefore not based on quantitative predictions of growth
and yield. In the absence of Norwegian guidelines, Swedish guidelines are applied in the south-
eastern part of the country (Glommen Skogeierforening 2005).
Basal area is most frequently used in the international literature and management practice to
measure and control stand density. The Norwegian tradition of using stem number has been
revised and replaced by basal area (Myklestad et al. 2006). Results from many thinning
experiments worldwide have demonstrated that the relationship between stand density, expressed
62
�
as basal area, and volume production follows an optimum curve. Volume increment can be
reduced at maximum basal area, but certainly decreases with decreasing basal area (Pretzsch 2001,
Zeide 2001, Skovsgaard & Vanclay 2008). The relationship varies according to species, site index
and age. Results from Finnish and Swedish thinning experiments on Norway spruce and Scots
pine confirm this relationship for the Fennoscandian boreal forest types (Eriksson 1990, Mäkinen
& Isomäki 2004a, 2004b, Agestam 2009). Most striking is the difference between spruce and pine:
spruce has a much larger flexibility to tolerate heavy thinnings without significant losses in
growth, whereas pine has already responded to moderate thinning intensities with significant
losses in growth as compared to growth at maximum stand density. Norwegian thinning experi-
ments have not been analysed in a similar way, but published results (Braastad & Tveite 2000,
2001) show similar trends.
Quantification of the relationship between basal area and growth at stand level and its variation
according to species, site and age for forests in Norway will be a valuable decision support tool for
forest management and will be available as the result of a current PhD project at the Norwegian
University of Life Sciences (Optimizing silviculture for bioenergy production in Norway, NFR
189098/I10). The shortcoming of the aforementioned tools is that they only produce predictions
about effects of thinning on total volume production, which is most relevant in a bioenergy
context, but not on diameter distribution. On the other hand, they are easier to establish and are
expected to be operational before individual-tree based growth models for Norway are available.
Thinning guidelines are much more urgently needed for Scots pine due to the current practice
whereby Scots pine stands are thinned much more often than Norway spruce stands, and due to
the fact that Scots pine reduces volume production significantly in response to heavy thinnings.
Thinning is currently carried out using harvesting machines. According to the most recent
guidelines (Glommen Skogeierforening 2005, Myklestad et al. 2006), basal area targets are set for
density after thinning. Stand density during operations is controlled by the subjective judgment of
the operator. Only rarely do operators leave their machines and check the resulting stand density
after thinning with relascope samples in order to adjust their activities. It can therefore be assumed
that stand density targets are frequently only met with low precision. As a consequence, a part of
the total yield might be lost if stands are thinned too heavily, or individual trees might not reach
the diameter targets if stands are thinned too moderately.
63
�
Ground-based laser scanning has been developed into a tool for forest inventorying (Hopkinson et
al. 2004, Parker et al. 2004, Thies et al. 2004, Watt & Donoghue 2005, Van der Zande et al. 2006).
Laser scanners can detect the position and diameter of individual trees in sample areas around the
scanner. Using a laser scanner mounted on a harvester would enable additional information to be
gained on stand density (number of trees, basal area) and the diameter distribution for the part of
the stand where the harvester operates. Such a tool will be available as a result of a current PhD
project at the Norwegian University of Life Sciences (Optimizing silviculture for bioenergy
production in Norway, NFR 189098/I10). The implementation of such a density management tool
will combine density and diameter distribution calculated from scanner data with density
management targets in order to indicate to the operator the number of trees to be removed and
their diameter.
Results from a series of spacing trials established in Norway spruce stands in Norway clearly
show the large effect that planting density has on growth and yield (Braastad 1970, 1979,
Haveraaen 1981, Handler 1988, 1990, Øyen et al. 2001). Niemistö (1995) reports similar trends
for birch in Finland. Gamborg (1997) reports on the effect of establishment density on woodchip
harvesting from Norway spruce stands in Denmark. Significant parts of the total yield can be lost
if stands are established with low stem numbers, either due to low planting density or due to
sparse natural regeneration. During the last 50 years the focus of forest management has shifted
from maximum volume production to high quality timber production. The latter also requires a
minimum density at establishment in order to control branch growth, but this density is much
lower than the minimum density required for maximum volume production. As a consequence,
experiments on establishment density have not been followed up in recent decades and results
have not been analysed or compiled. The increased focus on maximum volume production in the
context of bioenergy production means that those experiments again can be used as a valuable
data source. Compilation of published results, analysis of existing data, and remeasurement of
some of the experiments will allow the generation of decision support tools that will be available
as a result of a current PhD project at the Norwegian University of Life Sciences (Optimizing
silviculture for bioenergy production in Norway, NFR 189098/I10).
Pre-commercial thinning is the first opportunity to harvest wood for bioenergy and has significant
consequences for the total biomass production. Heikkilä et al. (2009) used a growth simulator to
analyse the effects of stand density after pre-commercial thinning and the time of the first thinning
for Norway spruce and Scots pine stands in Finland based on the assumption that biomass for
64
�
energy is harvested at the first thinning while industrial wood production is the main goal for the
remainder of the rotation. The stand density after pre-commercial thinning used in their scenarios
varied very little and consequently yield differences were minor, while the time of thinning
resulted in slightly larger differences in total yield. On the other hand, it is well known from
detailed analyses of extensive Swedish (Pettersson 1993) and Norwegian (Vestjordet 1977) pre-
commercial thinning experiments that density after pre-commercial thinning has a significant
effect on volume production up to the time of the first thinnings, i.e. trees removed during pre-
commercial thinnings cannot be harvested later for bioenergy purposes and consequently a major
proportion of the yield is lost. In their economic analysis, Heikkilä et al. (2009) also clearly
demonstrate that wood for bioenergy should be harvested during later thinnings. Despite
differences in management options it should be remembered that bioenergy harvesting under
energy prices is only profitable for the forest owner and the complete product chain, if subsidized
(Heikkilä et al. 2009).
In order to reduce the harvesting costs of bioenergy, geometric thinning has been proposed and the
operational consequences of this approach evaluated (Bergström et al. 2007, Spinelli et al. 2009).
It remains to analyse the consequences of this new type of geometric thinning for the growth and
quality of the remaining stand.
The main challenge for future research with regard to stand density management is the
development of growth models as decision support tools for forest managers. These new growth
models also have to include other biomass components than stem wood in order to be useful in a
bioenergy harvest context. It remains to be tested whether biomass expansion factors can be used
to dynamically predict the biomass components based on stem wood growth models under a
variety of different stand management options.
4.1.2. Use of pioneer species
Pioneer species are characterized by a rapid establishment after disturbances and fast growth in
their youth in order to occupy a given site and compete with other species. These traits are
sometimes used in forestry in order to maximize production. The two dominating tree species in
forestry in Norway today, Norway spruce and Scots pine, lack some typical pioneer traits,
especially rapid growth in their youth. Typical pioneers in the boreal forest types in Norway are
birch and other broadleaved species. They are often not used by the industry in Norway and are
therefore regarded as weeds and removed in early pre-commercial thinnings. Pioneer species can
65
�
be used in pure stands, often in short rotations in order to maximize volume production for fibre or
energy use. A review of the current knowledge about these systems is given in section 4.2.
Another method to take advantage of the rapid growth of pioneer species in traditional forestry is
to grow them in stratified mixtures in young stands.
Birch in stratified mixtures with Norway spruce in young forests is a resource that is currently not
used for bioenergy in Norway. By modifying current silvicultural practice and delaying pre-
commercial thinning until birch can be harvested for bioenergy it is not only possible to use this
resource but also to increase the total production of the stand.
In the boreal zone, Norway spruce forests established after clear-cutting are during the first years
dominated by spontaneous regeneration of pioneer tree species, mostly birch (Frivold 1986,
Hynynen et al. 2009). In order to concentrate production on the desired individuals, pre-
commercial thinnings largely remove birch. However, pre-commercial thinning is an expensive
intervention because no merchantable products are harvested. Even though a number of different
methods have been tested in recent decades, it has not been possible to further mechanize early
thinnings and therefore brush cutter is still the standard equipment. Due to their high costs, pre-
commercial thinnings are often dropped and only public subsidies guarantee a sufficient level of
pre-commercial thinning in Norway today. Trees removed during early thinning could be used for
bioenergy and we anticipate that this feedstock for bioenergy could be available at a price as low
as the harvesting costs. Forest owners might be willing to sell birch from early thinnings at very
low prices if stand tending is free of charge. Even though biomass harvesting costs in young
stands are currently much higher than for logging residues from final harvests (Hakkila 2005),
biomass from young stands contributes significantly to the woodchip harvest in Finland (Laitila
2008) and technical innovation might reduce the costs in the near future. Similar concepts for
energy wood thinnings have been presented for Finland (Hakkila 2005, Heikkilä et al. 2007, 2009)
and in Sweden (Johansson 2003).
By delaying early thinning until birch has reached larger dimensions, harvesting costs for biomass
can be reduced. Birch can, for example, reach 10 m in height in 20 years on typical Norway
spruce sites (Strand & Braastad 1967). Harvesting methods for trees of these dimensions are under
development (Jylhä & Laitila 2007, Spinelli et al. 2007, Laitila 2008) and we expect them to be
further developed in the near future due to the current demand for bioenergy in other Nordic
countries. By delaying the removal of birch shelters, the gowth of spruce can be reduced. It is
66
�
therefore important to quantify the growth dynamics of these two species when mixed in young
stands in order to optimize total production.
Precommercial thinning in the period 2004–2007 was reported to approximately 30,000 ha per
year and to 40,000 ha per year in 1996–1999 (http://www.ssb.no/emner/10/04/20/skogkultur/). For
the period 1995–2005, between 40,000 and 50,000 ha per year were finally cut, indicating that not
all of the young stands were thinned, even though due to the dominating clear cutting system it is
likely that almost the total area of final cuts needed some kind of pre-commercial thinning. The
numbers for harvestable birch biomass in young spruce and pine stands in Norway are lacking, but
assuming 50 m3/ha (Tham 1987, Tham 1994) and an area of 40,000 ha/year, a total of 2 million
m3/year might in theory be available for bioenergy purposes. For Finland, early thinning for
bioenergy has been estimated to have a potential of 4 million m3/year (Heikkilä et al. 2007).
To establish spruce stands, often only 1500–2000 trees ha-1 are planted in Norway today. This low
number of spruce seedlings does not utilize the biomass production potential of the site, mostly
because the species is characterized by slow growth in the early stage, as indicated by the differing
times for spruce and birch to reach breast height (Frivold 1982, Mielikäinen 1985). Planted spruce
on clear-cuts is mostly supplemented by spontaneous regeneration of spruce and pioneer trees,
mostly birch. Pioneer species establish and grow much faster in their youth than spruce and
therefore enable the site on which they grow to be put into full production after a much shorter
period than a pure spruce stand. When harvesting this extra production of biomass from pioneers
before they start retarding the growth of spruce, a spruce and birch mixture can produce more
biomass than a pure spruce stand (Tham 1994). Frivold & Frank (2002) investigated a number of
mixed stands of Norway spruce and birch in Norway and found that this mixture produces more
than pure spruce stands only up to a birch top height of 17 m, suggesting that only the different
growth dynamics of the two species is responsible for the additional production and that no
‘mixture effect’ (Kelty 1992, Frivold & Groven 1996, Frivold & Kolström 1999) is to be
expected. It is also important to notice that downy birch (Betula pubescens Ehrh.) is slower
growing than silver birch (Betula pendula Roth.) and therefore closer to Norway spruce in its
dynamics. Additional yield from a mixture of downy birch and Norway spruce can therefore be
less than the additional yield from silver birch, but this effect might be hidden by the fact that
downy birch is generally found on poorer sites than silver birch.
67
�
Braathe (1988) reported results from an experiment in Norway where the birch shelter had been
regulated to different densities. Norway spruce under the birch shelter was clearly affected by the
density of the shelter. An individual-tree based competition index could be used to explain the
reduction in height growth of spruce in response to competition. Based on the same experiment,
Gobakken & Næsset (2002) calibrated a growth model that predicts stand mean diameter growth
for Norway spruce based on, among other variables, a variable that describes competition between
spruce and birch. Brække & Granhus (2004) presented preliminary results from a number of pre-
commercial thinning experiments in mixed spruce and birch forests and modified the model
describing spruce height growth in response to birch density (Braathe 1988).
Based on a number of experiments in young mixed stands of spruce and birch in Sweden, an
individual-tree based growth model for this stand type was developed and scenarios were
presented for energy wood thinnings in Central Sweden (Tham 1988, 1989, 1994). The thinning
schedule included a pre-commercial thinning before age 15 to a density of 1200–3000 birches/ha,
an energy wood thinning at age 20, and final harvest of the remaining birches at age 30. At age 20
the standing volume of birch was 40–80 m3/ha, increasing with increasing birch density. For a
stand density of 1200 birches ha-1, birch produced 80 m3/ha to the age of 30 years, while
production of spruce was not reduced as compared to spruce without birch. Higher birch densities
reduced spruce production slightly, but the total yield was at the same level as that of the sparsest
birch shelter. The results also clearly indicate that individual tree dimensions of birch at the time
of the energy wood thinning will depend on the density of the birch shelter. The increasing yield
of birch in early stand phases with increasing initial stand density has also been reported for pure
birch stands in Finland (Niemistö 1995).
More recently, an individual-tree based growth model for young mixed stands in Southern Sweden
has been calibrated based on data from an extensive national survey of young stands (Fahlvik &
Nyström 2006). The model overpredicted birch height growth and underpredicted spruce diameter
growth when tested against data from experimental plots. Using this growth model, the effect of
different species mixtures and height stratifications was simulated and results indicated that only if
birches are higher than spruce will the total production of the mixed young stand be similar or
greater than that of a pure spruce stand (Fahlvik et al. 2005).
Gamborg (1997) showed how increasing stand density and mixture with pioneer species can in-
crease the woodchip harvest for Norway spruce stands in Denmark using growth models and
68
�
experimental data. Valkonen & Valsta (2001) demonstrated the economic benefits of growing a
birch and spruce mixture for Finnish conditions. Despite the fact that birch prices are much higher
in Finland and birches are in their simulations grown to timber dimensions, the sensitivity analysis
presented in this study indicates that even an earlier harvest of birch for energy might
overcompensate for the economic losses of reduced spruce production. Whole-tree harvesting of
birch in young mixed birch and spruce stands did not reduce growth in the following five-year
period when compared with a thinning without removal of birches in an experimental series in
Sweden (Mård 1998).
Birch shelters have also often been kept beyond the time indicated above for an energy wood
thinning. Even though birch admixtures have reduced the growth of spruce in a number of experi-
ments in Sweden (Mård 1996, Bergqvist 1999, Klang & Ekö 1999, Lindén 2003) and Norway
(Frivold 1982), the production of the birch shelter is often greater than the growth losses incurred
by the spruce understory. A birch shelter over spruce also increases the height differentiation of
spruce trees in the understory (Mård 1996). Growth models have been calibrated for older mixed
stands in Sweden (Agestam 1985, Ekö 1985), Estonia (Jogiste 1998, 2000), and Finland
(Mielikäinen 1980, 1985), and have been used to show the production of different mixtures of
spruce and birch.
Spruce trees released from a birch shelter have been studied in Sweden (Mård 1997a, 1997b),
where mechanical damage due to snow and wind affected up to 25% of the basal area, but no dif-
ferences in damages were recorded based on the season of release. It has also been reported from
Sweden that released spruce trees have faster height growth than previously unsuppressed spruce
trees (Tham 1988).
A birch shelter over young spruce saplings can also have a number of positive effects, such as
protecting againt microclimatic extremes, competing vegetation and voles (Bergan 1987) or re-
ducing waterlogging (Brække & Granhus 2004). Slower growth of spruce in the understorey
might also change wood quality, e.g. by reducing ring width and juvenile wood, improving wood
density, reducing stem taper, and reducing branch number and sizes. When studying spruce
planted at low density under birch in Sweden, smaller branches and ring width, lower numbers of
branches per whorl, fewer defects but equal wood density, stem taper, and total number of
branches per meter were found for sheltered than for open-grown spruce trees (Klang & Ekö
1999). In another experiment in northern Sweden, naturally regenerated spruce was grown under
69
�
a dense birch shelter up to a birch height of 13 m (spruce 2–4 m), and then the shelter was thinned
to a number of different densities (Bergqvist 1998, Bergqvist et al. 2000). Juvenile wood was
absent in the spruce trees that had been grown under the dense birch shelter up to a breast height
age of 10 years, but it is uncertain whether this was due to the suppressed growth or harsh climatic
conditions. The late modification of shelter density had little effects on wood density and fibre
properties. In a Finnish study of young stands consisting of pine planted at low density and
naturally regenerated birch (Valkonen & Ruuska 2003), the birch admixture significantly reduced
the branch diameter of pines, despite the fact that both tree species grow at similar height growth
rates in young stands.
Stratified mixtures of Norway spruce and birch in boreal Fennoscandinavia have been subject to a
lot of research. This stand type dominates after clear-cutting and potentially causes problems for
the main crop, which is considered to be Norway spruce, but offers a possibility to produce more
than a pure stand of Norway spruce. Even though this mixture has been considered under the
management objective to produce bioenergy from birch in Sweden and Finland earlier,
quantitative knowledge is lacking for Norwegian conditions. Conditions in Sweden, Finland and
Norway are similar, but application of Swedish and Finnish growth models and decision support
tools needs testing against data from Norway. An important part of the decision support for forest
managers is to quantify the competition between the birch shelter and the growth of Norway
spruce in order to optimize the time of the removal of the birch shelter. A master’s thesis at the
Norwegian University of Life Sciences is currently examining this topic.
4.1.3. Fertilization
Fertilization might be necessary to compensate for nutrient losses in the context of whole-tree
harvesting for bioenergy that might reduce site productivity (Egnell & Leijon 1997, 1999, Mård
1998, Jacobson et al. 2000, Egnell & Valinger 2003, Röser et al. 2008). The research project
‘Ecological consequences of increased biomass removal from forests in Norway’ includes an
investigation of the need for compensation of nutrient losses during whole-tree harvesting. Wood
ash recycling has frequently been tested as one measure to avoid nutrient export with whole-tree
harvesting (Röser et al. 2008). Nitrogen not only limits the growth of most forests in Norway but
is also missing in wood ash. High concentrations of toxic substances in wood ash pose a further
problem. Wood ash recycling is therefore currently of limited use in order to sustain or increase
production of biomass in forests.
70
�
On the other hand, fertilization is a well-known method to increase biomass production that has
been used in forestry (Ståhl 2009). Results from Scandinavian fertilization experiments
(Andersson et al. 1998, Ingerslev et al. 2001, Nilsen 2001, Nohrstedt 2001, Saarsalmi & Malkonen
2001, Högbom & Jacobson 2002) show that nitrogen fertilization is the most promising method
due to the fact that most of the forest ecosystems in central and northern Scandinavia are lacking
nitrogen. On the other hand, increasing nitrogen deposition will increase forest growth (Solberg et
al. 2004) and regionally saturate forest ecosystems with nitrogen, thereby reducing the effects of
nitrogen fertilization. Bergh et al. (2005), using a model-based analysis, demonstrated the effect of
fertilization on the growth of Norway spruce in Sweden when nutrient limitations are removed
completely. Biomass production can be drastically increased. Fertilization is therefore an im-
portant silvicultural method to increase biomass production.
Soil compaction and other effects of machine traffic as a consequence of intensified silviculture
might be a threat to sustainability. Fertilization and wood ash recycling are other operations that
can require additional machinery to be driven around in stands (Röser et al. 2008). It is therefore
of extreme importance to restrict the traffic of the machinery to strip roads, in addition to other
measures such as the use of specialized equipment and to restrict the use to periods with high
carrying capacity of the soil.
Fertilization of forests in Norway has only been practised on an insignificant amount of land and
almost ceased during the last few years (Nilsen 2001, Rognstad & Steinset 2008). Even though
there is ample evidence that nitrogen fertilization can increase production substantially, we are
missing decision support tools for forest managers who are predicting fertilization response based
on simple variables describing the sites in which they work. In practical forest management it is
unrealistic to base a decision about fertilization on analyses of the current nutritional status which
in turn is based on analyses of foliage and soil samples. Simpler tools need to be developed for
Norwegian conditions. For Sweden, similar decision support tools are available
(http://www.skogforsk.se/). Finally, it would be desirable to include fertilization responses into
growth models.
4.2. Energy forests
Pioneer tree species in short-rotation intensive silviculture are known to produce large amounts of
biomass, often much larger than those from traditional forestry. Energy forests have therefore
received much attention both in practice and research in many European countries after energy
71
�
supply problems came onto the international political agenda in the 1970s, but also in reaction to
agricultural overproduction. Sweden and Austria had large programmes to subsidize energy
forestry before they joined the European Union, because they had to solve agricultural
overproduction problems nationally. Consequently, a large number of research results are
available from Sweden, some of which might be applicable under Norwegian conditions. In
contrast to Sweden and Finland, energy forestry has never received much attention in Norway.
Apart from a single experiment (Hole 2001), no further research or practice is known to us.
Differences in growth conditions and ownership structure might explain some of the differences
between Norway and other Scandinavian countries when it comes to the practical application of
energy forestry, but the lack of public subsidies is definitely an important factor. Even if
agricultural land is unlikely to be available for energy in the near future, marginal lands (e.g.
power lines, roadsides, and borders to agricultural land) have already become a source of biomass
for energy and might be managed more intensively in the future. Energy forestry is currently in
focus in Denmark, both as a topic for large research programmes and practical applications.
A typical energy forest is established on agricultural land after soil preparation. Willows, aspen,
and poplars, often from selected individuals or genetically improved material, are the preferred
species. Rapid establishment by vegetative regeneration from stump sprouts is an essential factor
for the high productivity of such systems. In this respect, fertilization, vegetation control, pest and
disease control, and in some cases irrigation are essential for success. The net energy gain is
substantially reduced due to the high energy input of such intensive management compared to
traditional forestry.
Birch is not a species that is typically used for energy purposes because other pioneer species have
proven to be more productive if managed intensively. Nevertheless, with less intensive
management, birch in short rotation might be a viable option under Norwegian conditions, where
birch is the dominating pioneer in the natural forest type and often dominating in poorly managed
young stands. Therefore, research results related to birch in short rotation from other countries
(Ferm 1993, Johansson 1996, 1999b, 2007, Paukkonen & Kauppi 1998, Kaunisto 1999, Telenius
1999, Rydberg 2000, Hytonen & Luostarinen & Kauppi 2005, Walle et al. 2007a, 2007b) are of
great interest and should be tested under Norwegian conditions.
72
�
References Agestam, E. 1985. En produktionsmodell for blandskog av tall, gran og björk i Sverige. Swedish
University of Agricultural Sciences, Department of Forest Yield Research, Report, 15: 1–150.
Agestam, E. 2009. Gallring. Skogsskötselserien, 7: 1–83.
Albaugh, T.H., Bergh, J., Lundmark, T., Nilsson, U., Stape, J.L., Lee Allen, H. & Linder, S. 2009.
Do biological expansion factors adequately estimate stand-scale aboveground component
biomass for Norway spruce? Forest Ecology and Management 258: 2628–2637.
Andersson, F., Brække, F.H. & Hallbäcken, L. 1998. Nutrition and growth of Norway spruce
forest in a Nordic climatic and deposition gradient. TemaNord, 1998:566: 1–254.
Andreassen, K. 1994. Development and yield in selection forest. Medd. Skogforsk 47(5): 1–37.
Ask, J., Bergseng, E., Gobakken, T., Framstad, E. & Hoen, H.F. 2005. Effekter på økonomi og
skogstruktur ved skogbehandling tilpasset bevaring av biologisk mangfold i skog. I:
Virkemidler for forvaltning av biologisk mangfold - Delrapport 3: Tiltak og virkemidler for
vern av biodiversitet i skog og våtmarker. TemaNord 563: 155–206.
Asper, Å. 2007. Forskjellen mellom skogsvolum og salgsvolum i utvalgte granbestand. Master-
oppgave ved Institutt for naturforvaltning, Universitetet for miljø- og biovitenskap. 51 pp.
Athanassiadis, D., Melin, Y., Lundström, A. & Nordfjell, T. 2009. Marginalkostnader för skörd av
grot och stubbar från föryngringsavverkningar i Sverige. Arbetsrapport 261. Institutionen för
skoglig resurshushållning, Sveriges Lantbruksuniversitet. 25 pp. (In Swedish).
Bauger, E. 1995. Funksjoner og tabeller for kubering av stående trær. Furu, gran og sitkagran på
Vestlandet. Tree volume functions and tables. Scots pine, Norway spruce and Sitka spruce in
western Norway. Rapport fra Skogforsk 16/95. 26 pp.
Bergan, J. 1987. Virkningen av en bjørkeskjerm og vekst hos bartrær utplantet i Nord-Norge.
Norsk Institutt for skogforskning Rapport, 10/87: 1–47.
Bergh, J., Linder, S. & Bergström, J. 2005. Potential production of Norway spruce in Sweden.
Forest Ecology and Management 204(1): 1–10.
Bergqvist, G. 1998. Wood density traits in Norway spruce understorey: effects of growth rate and
birch shelterwood density. Annales Des Sciences Forestieres 55(7): 809–821.
Bergqvist, G. 1999. Wood volume yield and stand structure in Norway spruce understorey
depending on birch shelterwood density. Forest Ecology and Management 122: 221–229.
Bergqvist, G., Bergsten, U. & Ahlqvist, B. 2000. Fibre properties of Norway spruce of different
growth rates grown under birch shelterwoods of two densities. Canadian Journal of Forest
Research-Revue Canadienne De Recherche Forestiere 30(3): 487–494.
73
�
Bergseng, E., Ask, J.A., Framstad, E., Gobakken, T., Solberg, B. & Hoen, H.F. 2009. Biodiversity
protection and economics in long term boreal forest management – a detailed case for the
valuation of protection measures. Submitted to Forest Policy and Economics.
Bergström, D., Bergsten, U., Nordfjell, T. & Lundmark, T. 2007. Simulation of geometric
thinning systems and their time requirements for young forests. Silva Fennica 41(1): 137–147.
Berndes, G., Hoogwijk, M. & van den Broek, R. 2003. The contribution of biomass in the future
global energy supply: a review of 17 studies. Biomass and Bioenergy 25: 1–28.
Björheden, R., Gullberg, T. & Johansson, J. 2003. Systems analyses for harvesting small trees for
forest fuel in urban forestry. Biomass and Bioenergy 24: 389–400.
Blingsmo, K. 1984. Diametertilvekstfunksjoner for bjørk-, furu- og granbestand. Rapp. Nor.inst.
skogforsk. 7/84:1–22.
Blingsmo. K. 1985. Avsmalningsfunksjoner og -tabeller for bjørk (Taper functions and tables for
birch) Rapport fra Norsk institutt for skogforskning 10/85:1.35.
Blingsmo, K. & Veidahl, A. 1992. Funksjoner for bruttopris av gran- og furutrær på rot. Rapp.
Skogforsk. 8/92:1–23.
Bohlin, F. & Roos, A. 2002. Wood fuel supply as a function of forest owner preferences and
management styles. Biomass and Bioenergy 22:237–249.
Bolkesjø, T. 2004. Modelling supply, demand and trade in the Norwegian forest sector. Doctor
Scientiarum Thesis 2004:10, Agricultural University of Norway, Ås.
Bolkesjø, T., Trømborg, E. & Solberg, B. 2006. Bioenergy from the forest sector: Economic
potential and interactions with timber and forest products markets in Norway. Scandinavian
Journal of Forest Research 21: 175–185.
Bollandsås, O.M. 2007. Uneven-aged forestry in Norway: Inventory and management models.
PhD thesis, 2007:30. Department of Ecology and Natural Resource Management, Norwegian
University of Life Sciences, Ås.
Bollandsås, O.M., Hoen, H.F. & Lunnan, A. 2004a. Nullområder i skogbruket – en prinsipiell
betraktning. Rapport fra skogforskningen, 4/04: 35 s.
Bollandsås, O.M., Hoen, H.F. & Lunnan, A. 2004b. Nullområder i skogbruket – vurdering av
driftskostnader og miljøverdier. Rapport fra skogforskningen, 5/04: 23 s.
Bollandsås, O.M., Buongiorno, J. & Gobakken, T. 2008. Predicting the growth of stands of trees
of mixed species and size: A matrix model for Norway. Scandinavian Journal of Forest
Research 23: 167–178.
74
�
Bollandsås, O.M., Rekstad, I., Næsset, E. & Røsberg, I. 2009. Models for predicting aboveground
Biomass of Betula pubescens spp. czerepanóvii in mountain areas of southern Norway.
Scandinavian Journal of Forest Research: 24: 318–332.
Brantseg, A. 1967. Furu Sønnafjells. Kubering av stående skog. Funksjoner og tabeller. Volume
Functions and Tables for Scots Pine South Norway. Meddelelser fra Det Norske
Skogforsøksvesen 22: 689–739.
Brække, F.H. 1986. Distribution and yield of biomass from young Pinus sylvestris and Picea abies
stands on draines and fertilized peatland. Scandinavian Journal of Forest Research 1: 49–66.
Brække, F.H. & Granhus, A. 2004. Ungskogpleie i naturlig forynget gran på middels og høy
bonitet. Rapport fra skogforskningen, 10/04: 1–22.
Braastad, H. 1966. Volumtabeller for bjørk. Volume tables for Birch. Meddelelser fra Det Norske
Skogforsøksvesen 21: 23–78.
Braastad, H. 1970. Et forbandsforsøk i gran. Meddelelser fra det Norske Skogforsøksvesen, 28(5):
295–329.
Braastad, H. 1979. Vekst og stabilitet i et forbandsforsøk med gran. Meddelelser fra Norsk
institutt for skogforskning 34(7): 174–215.
Braastad, H. & Tveite, B. 2000. Tynning i granbestand. Effekten på tilvekst, dimensjonsfordeling
og økonomi. Rapport fra skogforskningen 4/00: 1–30.
Braastad, H. & Tveite, B. 2001. Tynning i gran- og furubestand. Effekt av tynning på
volumproduksjon, middeldiameter og diameter av de 800 grøvste trær per ha. Rapport fra
skogforskningen 10/01:1–27.
Braathe, P. 1988. Utviklingen av gjenvekst med ulike blandingsforhold mellom bartrær og løvtrær
- II. Norsk institutt for skogforskning Rapport 8/88: 1–50.
Bååth, H., Gällerspång, A., Hallsby, G., Lundström, A., Löfgren, P., Nilsson, M. & Ståhl, G.
2002. Remote sensing, field survey, and long-term forecasting: an efficient combination for
local assessments of forest fuels. Biomass and Bioenergy 22(3): 145–157.
Cochran, W.G. 1977. Sampling techniques. 3rd Edition. New York, Wiley.
Dale, Ø. & Stamm, J. 1994. Grunnlagsdata for kostnadsanalyse av alternative hogstformer
Rapp.Skogforsk 7/94: 1–37.
Dale, Ø., Kjøstelsen, L. & Aamodt, H.E. 1993. Mekaniserte lukkede hogster. Side 3–23 i:
Aamodt, H.E. (ed.). Flerbruksrettet driftsteknikk. Rapp.Skogforsk 20/93:1–40.
Doruska, P., Patterson, D. & Posey, T. 2004. Stand-Level green biomass equations for sawtimber-
sized loblolly pine stands. Journal of the Arkansas Academy of Science 58:46–51.
Egnell, G. 2009. Skogsbränsle. Skogsskötselserien, 17: 1–68.
75
�
Egnell, G. & Leijon, B. 1997. Effects of different levels of biomass removal in thinning on short-
term production of Pinus sylvestris and Picea abies stands. Scandinavian Journal of Forest
Research 12(1): 17–26.
Egnell, G. & Leijon, B. 1999. Survival and growth of planted seedlings of Pinus sylvestris and
Picea abies after different levels of biomass removal in clear-felling. Scandinavian Journal of
Forest Research 14(4): 303–311.
Egnell, G. & Valinger, E. 2003. Survival, growth, and growth allocation of planted Scots pine
trees after different levels of biomass removal in clear-felling. Forest Ecology and
Management 177(1-3): 65–74.
Eid, T. 1998. Langsiktige prognoser og bruk av prestasjonsfunksjoner for å estimere kostnader
ved mekanisert drift. Rapport fra skogforskningen 7/98: 1–31.
Eid, T. 2004. Testing a large-scale forestry scenario model by means of successive inventories on
a forest property. Silva Fennica 3: 305–317.
Eid, T. 2007. Utvikling og bruk av prognoseverktøy for beslutningsstøtte i skogforvaltning.
(Development and use of forestry scenario models for decision-support in forest management)
INA-Fagrapport nr. 10. 61 p. (In Norwegian with English summary).
Eid, T. & Hobbelstad, K. 1999. AVVIRK-2000 - et Edb-program for langsiktige investerings-,
avvirknings- og inntektsprognoser i skog. Rapport Supplement fra skogforskningen 8/99: 1–
63.
Eid, T. & Hobbelstad, K. 2000. AVVIRK-2000 - a large scale scenario model for long-term
investment, income and harvest analyses. Scandinavian Journal of Forest Research 15: 472–
482.
Eid, T. & Hobbelstad, K. 2005. Langsiktige investerings-, avvirknings- og inntektsanalyser for
skog med Avvirk-2000. Aktuelt fra skogforskningen.2/05: 1–29.
Eid, T. & Hobbelstad, K. 2006. Langsiktige konsekvensanalyser – etterprøving basert på
Landsskogtakseringens prøveflater og avvirkningsstatistikk. Rapport fra skogforskningen.
2/06: 1–26.
Eid, T. & Tuhus, E. 2001. Models for individual tree mortality in Norway. Forest Ecology and
Management 154: 69–84.
Eid, T. & Øyen, B.-H. 2003. Models for prediction of mortality in even-aged forest. Scandinavian
Journal of Forest Research 18: 64–77.
Eid, T., Hoen, H.F. & Økseter, P. 2001. Economic consequences of sustainable forest
management regimes at non-industrial forest owner level in Norway. Forest Policy and
Economics 2: 213–228.
76
�
Eid, T., Hoen, H.F. & Økseter, P. 2002. Timber production possibilities of the Norwegian forest
area and measures for a sustainable forestry. Forest Policy and Economics 4: 187–200.
Eide, E. & Langsæter, A. 1954. Avsmalningstabell for granskog. Utvidet og revidert utgave.
Meddelelser fra Det norske Skogforsøksvesen 3 (12): 1–62 s.
Ekö, P.M. 1985. En produktionsmodell för skog i Sverige, baserad på bestand från
riksskogtaxeringens provytor. Sveriges Lantbruksuniversitetet, Institutionen for skogsskötsel,
rapporter, 16: 1–224.
Eriksen, R., Hobbelstad, K. & Aalde, H. 2004. Skogbruk og inngrepsfrie naturområder. En
analyse av sammenhengen mellom tilgjengelighet til skogressursene, bygging av skogsveier
og bevaring av inngrepsfrie naturområder. NIJOS-rapport 11/2004. 54 pp.
Eriksson, H. 1990. Hur har det gått med høggallringen? Sveriges Skogvårdsforbunds Tidskrift, 2:
43–57.
Eriksson, H., Eklundh, L., Kuusk, A. & Nilson, T. 2006. Impact of understory vegetation on forest
canopy reflectance and remotely sensed LAI estimates. Remote Sensing of Environment 103,
pp. 408–418.
Eriksson, L.E.B., Magnusson, M., Fransson, J.E.S., Sandberg, G. & Ulander, L.M.H. 2007. Stem
volume estimation for boreal forest using ALOS PALSAR. 5th International Symposium on
Retrieval of Bio- and Geophysical Parameters from SAR Data for Land Applications, Bari,
Italy, September 25–28, 2007, ESA, Istituto di Studi sui Sistemi Intelligenti per
l´Automazione. CD.
Fahlvik, N., Agestam, E., Nilsson, U. & Nyström, K. 2005. Simulating the influence of initial
stand structure on the development of young mixtures of Norway spruce and birch. Forest
Ecology and Management 213(1-3): 297.
Fahlvik, N. & Nyström, K. 2006. Models for predicting individual tree height increment and tree
diameter in young stands in southern Sweden. Scandinavian Journal of Forest Research 21:
16–28.
FAO. 2007. Global Forest Resources Assessment 2010. Specification of National Reporting
Tables for FRA 2010. Forest Resources Assessment Program. Working Paper 135. Rome. 51
pp.
Ferm, A. 1993. Birch Production andUtilization For Energy. Biomass and Bioenergy, 4(6): 391–
404.
Frivold, L.H. 1982. Bestandsstruktur og produksjon i blandingsskog av bjørk (Betula verrucosa
Ehrh., B. pubescens Ehrh.) og gran (Picea abies (L.) Karst.) i Sydøst-Norge. Meldinger fra
Norges landbrukshøgskole, 61(18): 1–108.
77
�
Frivold, L.H. 1986. Frøforyngelse av bjørk og gran på hogstflater i lavlandet østafjells, i relasjon
til vegetasjonstype og fuktighet. Meddelelser fra Norsk Institutt for Skogforskning, 39: 67–84.
Frivold, L.H. & Frank, J. 2002. Growth of mixed birch-coniferous stands in relation to pure
coniferous stands at similar sites in south-eastern Norway. Scandinavian Journal of Forest
Research 17(2): 139–149.
Frivold, L.H. & Groven, R. 1996. Yield and management of mixed stands of spruce, birch and
aspen. Norwegian Journal of Agricultural Sciences, supp. no. 24: 1–21.
Frivold, L.H. & Kolström, T. 1999. Yield and treatment of mixed stands of boreal tree species in
Fennoscandia. In: Oolsthorn, A.F.M. et al. Management of mixed-species forest: silviculture
and economics, IBN-DLO, Wageningen, The Netherlands. Scientific contributions, 15: 37–
45.
Gamborg, C. 1997. Maximising the production of fuelwood in different silvicultural systems.
Biomass and Bioenergy 13(1-2): 75–81.
Gjertsen, A.K. 2007. Accuracy of forest mapping based on Landsat TM data and a kNN-based
method. Remote Sensing of Environment 110: 420–430.
Gjølsjø, S. & Belbo, H. 2008. Skogbrensel i Norge – Status og muligheter. Oppdragsrapport fra
Skog og Landskap 23/2008. 28 pp.
Gjølsjø, S. & Hobbelstad. K. 2009. Energipotensialet fra skogen i Norge. Oppdragsrapport fra
Skog og Landskap 09/2009. 11 pp.
Glommen Skogeierforening. 2005. Tynningsdiagrammer i Glommen Skogeierforening. Glommen
Skogeierforening, Elverum, 12 pp.
Gobakken, T. 2003. Brukerveiledning til SGIS- et skoglig geografisk informasjonssystem.
Versjon 2.1. 26 pp. Upublisert brukermanual.
Gobakken, T. 2007. Brukerveiledning til dataprogrammet T(re). Versjon 1.0. 60 pp. Upublisert
brukermanual.
Gobakken, T. & Næsset, E. 2002. Spruce diameter growth in young mixed stands of Norway
spruce (Picea abies (L.) Karst.) and birch (Betula pendula Roth B-pubescens Ehrh.). Forest
Ecology and Management, 171(3): 297–308.
Gobakken, T., Lexerød, N. & Eid, T. 2008. T – A forest simulator for bioeconomic analyses based
on models for individual trees. Scandinavian Journal of Forest Research 23: 250–265.
Gulbrandsen, K. 1980. Taksering av hogstavfall. Meddr Nor. inst. skogforsk. 35: 361–381.
Gulbrandsen, K. 1981. Forholdet mellom skogsvolum og tømmervolum for gran. Rapp.Nor. inst.
skogforsk. 8/81: 1–20.
78
�
Hakkila, P. 2005. Fuel from early thinnings. International Journal of Forest Engineering 16(1):
11–14.
Handler, M.M. 1988. Forbandtsforsøg i granskov. Tæthed, diameterspredning og kvalitet: Forsøg
928, Mathiesen-Eidsvold Værk, Hurdal. Rapport fra Norsk institutt for skogforskning 1/88:
1–20.
Handler, M.M. 1990. Et planteavstandsforsøg med gran, Picea abies (L.) Karst.: Forsøg 626/627,
Spikkestad. Meddelelse fra Norsk institutt for skogforskning 43(4): 1–75.
Haveraaen, O. 1981. Vekstutvikling i et 20-årig forbandsforsøk med gran, Picea abies (L.) Karst.
Meldinger fra Norges Landbrukshøgskole 60 (13): 1–12.
Heikkilä, J., Sirén, M., Ahtikoski, A., Hynynen, J., Sauvula, T. & Lehtonen, M. 2009. Energy
Wood Thinning as a Part of the Stand Management of Scots Pine and Norway Spruce. Silva
Fennica 43(1): 129–146.
Heikkilä, J., Sirén, M. & Äijälä, J.O. 2007. Management alternatives of energy wood thinning
stands. Biomass and Bioenergy 31(5): 255–266.
Henning, J.G. & Radtke, P.J. 2006a. Detailed stem measurements of standing trees from ground-
based scanning lidar. Forest Science 52: 67–80.
Henning, J.G. & Radtke, P.J. 2006b. Ground-based laser imaging for assessing three-dimensional
forest canopy structure. Photogrammetric Engineering and Remote Sensing 72: 1349–1358.
Hobbelstad, K. 2007a. Ressurssituasjonen i Hedmark og Oppland. Oppdragsrapport fra skog og
Landskap 13/2007: 1–13.
Hobbelstad, K. 2007b. Skogressurser i Sør-Østerdal. Oppdragsrapport fra skog og Landskap
14/2007: 1–13.
Hoen, H.F. 1996. Forestry scenario modelling for economic analysis – experiences using the
GAYA-JLP model, ed. Pavinen, R., Roihuvuo, L. & Siitonen, M. Vol. 5, pp. 217–228.
Hoen, H.F. & Eid, T. 1990. En modell for analyse av behandlingsstrategier for en skog ved
bestandssimulering og lineær programmering. Rapport fra Norsk institutt for skogforskning
9/90: 1–35. (In Norwegian with English summary.)
Hoen, H.F. & Gobakken, T. 1997. Brukermanual for bestandssimulatoren GAYA v.1.20.
Upublisert brukermanual, Institutt for skogfag, NLH, Ås.
Hoen, H.F. & Solberg, B. 1994. Potential and economic efficiency of carbon sequestration in
forest biomass through silvicultural management. Forest Science 40: 429–451.
Hoen, H.F. & Solberg, B. 1999. Policy options in carbon sequestration via sustainable forest
management: an example from the north, ed. M. Palo, Vol. 15, The United Nations University,
World Institute for Development Economics Research, pp. 117–132.
79
�
Hoen, H.F., Eid, T. & Økseter, P. 2001. Timber production possibilities and capital yields from
the Norwegian forest area. Silva Fennica 3: 249–264.
Hoen, H.F., Eid, T., Veisten, K. & Økseter, P. 1998. Økonomiske konsekvenser av tiltak for et
bærekraftig skogbruk. Forutsetninger og metodebeskrivelse. Rapport Supplement fra
skogforskningen 6/98: 1–48.
Hole, E.E. 2001. Bioenergi. Miljø, teknikk og marked. Energigården, 390 pp.
Hopkinson, C., Chasmer, L., Young-Pow, C. & Treitz, P. 2004. Assessing forest metrics with a
ground-based scanning lidar. Canadian Journal of Forest Research 34(3): 573–583.
Huang, P., Seifert, S., Wipfler, P. & Pretzsch, H. 2007. Determination of canopy gap dynamic of
Norway spruce and European beech after thinning – a new approach based on terrestrial laser
scanning. European Journal of Forest Research.
Hynynen, J., Niemistö, P., Viherä-aarnio, A., Brunner, A., Hein, S. & Velling, P. 2009.
Silviculture of birch (Betula pendula Roth & Betula pubescens Ehrh.) in northern Europe.
Forestry, in print.
Hytonen, J. & Kaunisto, S. 1999. Effect of fertilization on the biomass production of coppiced
mixed birch and willow stands on a cut-away peatland. Biomass and Bioenergy 17(6): 455–
469.
Högbom, L. & Jacobson, S. 2002. Kväve 2002 - en konsekvensbeskrivning av skogsgödslingen i
Sverige. Skogforsk Redogörelse, 6: 1–42.
Ingerslev, M., Mälkönen, E., Nilsen, P., Nohrstedt, H.-Ö., Óskarsson, H. & Raulund-Rasmussen,
K. 2001. Main findings and future challenges in forest nutritional research and management
in the Nordic countries. Scandinavian Journal of Forest Research 16(6): 488–501.
Jalkanen, A., Mäkipää, R., Ståhl, G., Lehtonen, A. & Petersson, H. 2005. Estimation of biomass
stock of trees in Sweden: comparison of biomass equations and age-dependent biomass
expansion factors. Annals of Forest Science 62: 845–851.
Jacobson, S., Kukkola, M., Malkonen, E. & Tveite, B. 2000. Impact of whole-tree harvesting and
compensatory fertilization on growth of coniferous thinning stands. Forest Ecology and
Management 129(1-3): 41–51.
Jogiste, K. 1998. Productivity of mixed stands of Norway spruce and birch affected by population
dynamics: a model analysis. Ecological Modelling 106: 77–91.
Jogiste, K. 2000. A basal area increment model for Norway spruce in mixed stands in Estonia.
Scandinavian Journal of Forest Research 15(1): 97–102.
Johansson, T. 1996. Management of birch forest. Norwegian Journal of Agricultural Sciences,
suppl. 24: 7–20.
80
�
Johansson, T. 1999a. Biomass production of Norway spruce (Picea abies (L.) Karst.) growing on
abandoned farmland. Silva Fennica 33 (4): 261–280.
Johansson, T. 1999b. Biomass equations for determining fractions of European aspen growing on
abandoned farmland and some practical implications. Biomass and Bioenergy 17: 471–480.
Johansson, T. 2003. Mixed stands in Nordic countries – a challenge for the future. Biomass and
Bioenergy 24(4-5): 365–372.
Johansson, T. 2007. Biomass production and allometric above- and below-ground relations for
young birch stands planted at four spacings on abandoned farmland. Forestry 80(1): 41–52.
Johnsen, P.H. 2009. Above ground biomass of Sitka spruce (Picea sitchensis) in Norway. MSc
thesis, Norwegian University of Life Sciences (In Norwegian).
Junginger, M., Faaij, A.P.C., Bjorheden, R. & Turkenburg, W.C. 2005. Technological learning
and cost reductions in wood fuel supply chains in Sweden. Biomass and Bioenergy 29 (6),
399–418.
Junginger, M., de Visser, E., Hjort-Gregersen, K., Koorneef, J., Raven, R., Faaij, A. &
Turkenburg, W. 2006. Technological learning in bioenergy systems. Energy Policy 34, 4024–
4041.
Jylhä, P. & Laitila, J. 2007. Energy wood and pulpwood harvesting from young stands using a
prototype whole-tree bundler. Silva Fennica 41(4): 763–779.
Kairi�kštis, L. & Jaskelevi�ius, B. 2003. Forest Energy Resources and their Utilization in
Lithuania. Baltic Forestry 9 (2): 29–41
KanEnergi . 2003, Bioenergiressurser i Norge. Utarbeidet av Lena N. Berg, Per F. Jørgensen,
Petter Heyerdahl og Gunnar Wilhelmsen for Norges vassdrags- og energidirektorat, Oslo.
KanEnergi. 2007. Biomasse – nok til alle formål. Utredning om dagens potensial,
betalingsvillighet og mulige effekter for annen industri ved bygging av kraftvarmeanlegg
basert på bioenergi. KanEnergi, mars 2007.
Karjalainen, T., Asikainen, A., Ilavský, J., Zamboni, R., Hotari, K.-E. & Röser, D. 2004.
Estimation of energy wood potential in Europe. Working Papers of the Finnish Forest
Research Institute 6. 43 p.
Karlsson, J. 2007. Produktivitet vid stubblyftning (Productivity at Stump Lifting). Arbetsrapport
168. Institutionen för skoglig resurshushållning, Sveriges Lantbruksuniversitet. 52 pp. (In
Swedish with English summary).
Kelty, M.J. 1992. Comparative productivity of monocultures and mixed-species stands. In: Kelty,
M.J. et al. The ecology and silviculture of mixed-species forests, Kluwer Academic
Publishers: 125–141.
81
�
Kjelvik, S. 1974. Primærproduksjon i en subalpin bjørkeskog på Maurset i Øvre-Eidfjord,
Hordaland [Primary production in a sub-alpine birch forest in Maurset in Øvre Eidfjord
municipality,Hordaland County]. Doctoral dissertation, Norwegian Agricultural University,
1974). Licentiate thesis, NLH, Ås. (In Norwegian with English summary.)
Klang, F. & Ekö, P.M. 1999. Tree properties and yield of Picea abies planted in shelterwoods.
Scandinavian Journal of Forest Research 14(3): 262–269.
Korsmo, H. 1995. Weight equations for determination of biomass fractions of young hardwoods
from natural regenerated stands. Scandinavian Journal of Forest Research 10: 333–346.
Kärhä, K. 2006. Whole-tree harvesting in young stands in Finland. Forestry Studies 45: 118–134.
Kärhä, K. & Vartiamäki, T. 2006. Productivity and costs of slash bundling in Nordic conditions.
Biomass and Bioenergy 30: 1043–1052.
Kärhä, K., Jouhiaho, A., Mutikainen, A. & Mattila, S. 2005. Mechanized energy wood harvesting
from early thinning. International Journal of Forest Engineering 16: 15–25.
Kärkkäinen, L. 2005. Evaluation of performance of tree-level biomass models for forestry
modeling and analyses. Finnish Forest Research Institute, Research Papers 940. 123 pp.
Ladanai, S. & Vinterbäck, J. 2009 Global Potential of Sustainable Biomass for Energy. Swedish
University of Agricultural Sciences, Department of Energy and Technology. Report. 29 pp.
Laitila, J. 2006. Cost and sensitive analysis tools for forest energy procurement chains. Forestry
Studies 45: 5–10.
Laitila, J. 2008. Harvesting technology and the cost of fuel chips from early thinning. Silva
Fennica 42: 267–283.
Laitila, J. & Asikainen, A. 2004. Productivity and costs of stump wood harvesting. Confidential
report to UPM Forest. 33 pp.
Laitila, J., Asikainen, A. & Nuutinen, Y. 2007. Forwarding of whole trees after manual and
mechanized felling bunching in pre-commercial thinnings. International Journal of Forest
Engineering 18: 29–39.
Laitila, J., Kärhä, K. & Jylhä, P. 2009. Time consumption models and parameters for off- and on-
road transportation of whole-tree bundles. Baltic Forestry 15: 105–114.
Langerud, B, Størdal, S., Wiig, H. & Ørbeck, M. 2007. Bioenergi i Norge – potensialer, markeder
og virkemidler. ØF-rapport nr. 17/2007. 198 pp.
Lappi, J. 2005. J-user’s guide. Finnish For. Res. Inst. Sounenjoki. Version 0.9.4 February, 2005.
Larsson, J.Y. & Hylen, G. 2007. Statistikk over skogforhold og skogressurser i Norge registrert I
perioden 2000-2004. Statistics of Forest Conditions and Forest Resources in Norway. Viten
fra Skog og landskap 1/07. 91 pp.
82
�
Lattimore, B., Smith, C.T., Titus, B.D., Stupak, I. & Egnell, G. 2009. Environmental factors in
woodfuel production: Opportunities, risks, and criteria and indicators for sustainable
practices. Biomass and Bioenergy 33: 1321–1342.
Lauren, A., Sikanen, L., Asikainen, A., Koivusalo, H., Palviainen, M., Kokkonen, T., Kellomaki,
S. & Finer, L. 2008. Impacts of logging residue and stump removal on nitrogen export to a
stream: A modelling approach. Scandinavian Journal of Forest Research 23: 227–235.
Lazdi�š, A., von Hofsten, H., Dagnija, L. & Lazd�ns, V. 2009. Productivity and costs of stump
harvesting for bioenergy production in Latvian conditions. Engineering for rural development.
Jelgava, 28.-29.05.2009, 194–201.
Lehtonen, A., Makipaa, R., Heikkinen, J., Sievanen, R. & Liski, J. 2004. Biomass expansion
factors (BEFs) for Scots pine, Norway spruce and birch according to stand age for boreal
forests. Forest Ecology and Management 188: 211–224.
Levy, P.E., Hale, S.E. & Nicoll, B.C. 2004. Biomass expansion factors and root: shoot ratios for
coniferous tree species in Great Britain. Forestry 77(5), 421–430.
Lexerød, N. 2005. Recruitment models for different tree species in Norway. Forest Ecology and
Management 206: 91–108.
Lexerød, N. 2007. Planning, management and economy of selective cutting in Norway. PhD thesis
2007:30. Department of Ecology and Natural Resource Management, Norwegian University
of Life Sciences, Ås.
Lexerød, N. & Eid, T. 2005. Recruitment models for Norway spruce, Scots pine, birch and other
broadleaves in young even-aged forests in Norway. Silva Fennica 39: 91–106.
Lexerød, N. & Eid, T. 2006. Assessing suitability for selective cutting using a stand level index.
Forest Ecology and Management 237:503-512.
Lexerød, N.L. & Gobakken, T. 2002. Normerte prisfunksjoner for grantrær med og uten rotråte
[Standardized price functions for spruce with and without root rot]. In B.-H. Øyen (Ed.),
Modellering av skogproduksjon for økologisk og økonomisk forvaltning. Aktuelt fra
skogforskningen 6/02: 15–19. (In Norwegian.)
Lexerød, N., Gobakken, T. & Eid, T. 2009. Economic efficiency of selective cutting under
different timber price scenarios. Manuscript.
Lindhe, A. 2009. Stubbarnas biologiska betydelse underskattas. Potentiella konsekvenser av
storskalig stubbskörd for den vedberoende biologisk mangfalden. [Biological importance of
stumps is underestimated. Potential consequences of large-scale stump harvesting for the
wood-dependent biodiversity.] WWF rapport. Världsnaturfonden, Solna, Sverige, 16 pp.
83
�
Lindén, M. 2003. Increment and yield in mixed stands with Norway spruce in southern Sweden.
Acta universitatis agriculturae sueciae, silvestria, 260.
Living Forest. 1998. Standardutredninger fra Levende Skog. Levende Skogs rapport 9a-d. Oslo.
508 p.
Living Forest. 2001. Presisering av Levende Skogs standarder. Levende Skog, Oslo. 10 pp.
Living Forest. 2010. Standard for Sustainable Forest Management in Norway. 39 pp.
http://www.levendeskog.no/levendeskog/vedlegg/08Levende_Skog_standard_Bokmaal.pdf
Long, J.N., Dean, T.J. & Roberts, S.D. 2004. Linkages between silviculture and ecology:
examination of several important conceptual models. Forest Ecology and Management 200:
249–261.
Luostarinen, K. & Kauppi, A. 2005. Effects of coppicing on the root and stump carbohydrate
dynamics in birches. New Forests 29(3): 289–303.
Luther, J.E., Fournier, R.A,. Piercey, D.E., Guindon, L. & Hall, R.J. 2006. Biomass mapping
using forest type and structure derived from Landsat TM imagery. International Journal of
Applied Earth Observations and Geoinformation 8: 173–187.
Magnusson, M. & Fransson, J.E.S. 2004. Combining airborne CARABAS-II VHF SAR data and
optical SPOT-4 satellite data for estimation of forest stem volume. Canadian Journal of
Remote Sensing 30: 661–670.
Malinen, J., Pesonen, M., Määttä, T. & Kajanus, M. 2001. Potential harvest for wood fuels
(energy wood) from logging residues and first thinning in Southern Finland. Biomass and
Bioenergy 20:189–96.
Marklund, L.G. 1987. Biomass functions for Norway spruce (Picea abies (L) Karst.) in Sweden
(Rapport 43). Department of Forest Survey, Swedish University of Agricultural Sciences.
Marklund, L.G. 1988. Biomassafunktioner för tall, gran och björk i Sverige. Sveriges
Landbruksuniversitet. Rapporter Skog, 45: 1–73.
MCPFE. 2007. State of Europe’s Forests 2007: The MCPFE Report on Sustainable Forest
Management in Europe. Ministerial Conference on the Protection of Forests in Europe. 247
pp.
Mielikäinen, K. 1980. Mänty-koivusekametsiköiden rakenne ja kehitys. Summary: Structure and
development of mixed pine and birch stands. Communicationes instituti forestalis fenniae
99(3): 1–82.
Mielikäinen, K. 1985. Koivusekoituksen vaikutus kuusikon rakenteeseen ja kehitykseen.
Summary: Effect of an admixture of birch on the structure and development of Norway
spruce stands. Communicationes instituti forestalis fenniae 133: 1–79.
84
�
Molteberg, D. & Høibø, O. 2007. Modelling of wood density and fibre dimensions in mature
Norway spruce. Canadian Journal of Forest Research 37(8): 1373–1389.
Muukkonen, P. & Makipaa, R. 2006. Biomass equations for European trees: Addendum. Silva
Fennica 40: 763–773.
Myklestad, G., Follum, J.-R., Huse, K.J., Johnsrud, T.-E. & Pettersen, J. 2006. Tynning i gran- og
furuskog. Skogbrukets kursinstitutt, Biri. 59 pp.
Mäkinen, H. & Isomäki, A. 2004a. Thinning intensity and growth of Norway spruce stands in
Finland. Forestry 77(4): 349–364.
Mäkinen, H. & Isomäki, A. 2004b. Thinning intensity and growth of Scots pine stands in Finland.
Forest Ecology and Management 201: 311–325.
Mård, H. 1996. The influence of a birch shelter (Betula spp) on the growth of young stands of
Picea abies. Scandinavian Journal of Forest Research 11: 343–350.
Mård, H. 1997a. Damage and growth response in suppressed Picea abies after removal of
overstorey birch (Betula spp.). Scandinavian Journal of Forest Research 12(3): 248–255.
Mård, H. 1997b. Stratified mixture of young Norway spruce and birch as an alternative to pures
stands of Norway spruce. Acta universitatis agriculturae sueciae 35.
Mård, H. 1998. Short-term growth effects of whole-tree harvest in early thinnings of Birch (Betula
spp) and Picea abies. Scandinavian Journal of Forest Research 13(3): 317–323.
Niemistö, P. 1995. Influence of Initial Spacing and Row-To-Row Distance on The Growth and
Yield of Silver Birch (Betula-Pendula). Scandinavian Journal of Forest Research 10(3): 245–
255.
NIJOS & NORSKOG. 1999. Klargjøring av avvirkningsmuligheter i norsk skogbruk. Norsk
Institutt for Jord- og Skogkartlegging. Rapport 10/99. 52 s.
Nilsen, P. 2001. Fertilization experiments on forest mineral soils: A review of the Norwegian
results. Scandinavian Journal of Forest Research 16(6): 541–554.
Nilson, T., Kuusk, A., Lang, M. & Lükk, T. 2003. Forest Reflectance Modeling: Theoretical
Aspects and Applications. Ambio 32, pp. 535–541.
Nohrstedt, H.O. 2001. Response of coniferous forest ecosystems on mineral soils to nutrient
additions: A review of Swedish experiences. Scandinavian Journal of Forest Research 16(6):
555–573.
Nord-Larsen, T. & Talbot, B. 2004. Assessment of forest-fuel resources in Denmark: technical
and economic availability. Biomass and Bioenergy 27: 97–109.
NOU 1998. Energi- og kraftbalansen mot 2020. NOU 1998: 11. Olje- og energidepartementet 3.
juli 1998.
85
�
Nurmi, J. 2007. Recovery of logging residues for energy from spruce (Picea abies) dominated
stand. Biomass and Bioenergy 31:375–380.
Næsset, E. 1997. A spatial decision support system for long-term forest management planning by
means of linear programming and a geographical information system. Scandinavian Journal
of Forest Research 12: 77-88.
Næsset, E. 2007. Airborne laser scanning as a method in operational forest inventory: Status of
accuracy assessments accomplished in Scandinavia. Scandinavian Journal of Forest Research
22: 433–442.
Næsset, E. & Gobakken, T. 2008. Estimation of above- and below-ground biomass across regions
of the boreal forest zone using airborne laser. Remote Sensing of Environment 112(6): 3079–
3090.
Næsset, E., Gobakken, T. & Nelson, R. 2009. Sampling and mapping forest volume and biomass
using airborne LIDARs. Proceedings of the Eight Annual Forest Inventory and Analysis
Symposium, October 16–19, 2006, Monterey, CA, USA, pp. 297–301.
OED. 1997. Det interdepartementale arbeidsutvalget for bioenergi. Olje- og energidepartementet,
Oslo. http://www.regjeringen.no/nb/dep/oed/dok/rapporter_planer/rapporter/1997/rapport-fra-
det-interdepartementale-arbe.html?id=105239.
Olander, L.P., Gibbs, H.K., Steininger, M., Swenson, J.J. & Murray, B.C. 2008. Reference
scenarios for deforestation and forest degradation in support of REDD: a review of data and
methods. Environ. Res. Letters 3: 1–12.
Opdahl, H. 1987. Bjørkeskog til energiformål i Nord-Østerdalen [Birch forests for energy
purposes in the north of Østerdalen]. Norsk Skogbruk 2, 28–30. (In Norwegian).
Ovaskainen, H., Palander, T., Jauhiainen, M., Lehtimäki, J., Tikkanen, L. & Nurmi, J. 2008.
Productivity of energywood harvesting chain in different stand conditions of early thinnings.
Baltic Forestry 14: 149–154.
Padari, A., Muiste, O., Mitt, R. & Pärn, L. 2009. Estimation of Estonian wood fuel resources.
Baltic Forestry: 77–85.
Palander, T., Vesa, L., Tokola, T., Pihlaja, P. & Ovaskainen, H. 2009. Modelling the stump
biomass of stands for energy production using a harvester data management system.
Biosystems Engineering: 102: 69–74.
Parker, G.G., Harding, D.J. & Berger, M.L. 2004. A portable LIDAR system for rapid
determination of forest canopy structure. Journal of Applied Ecology, 41(4): 755–767.
86
�
Paukkonen, K. & Kauppi, A. 1998. Effect of coppicing on root system morphology and its
significance for subsequent shoot regeneration of birches. Canadian Journal of Forest
Research 28(12): 1870–1878.
Penman, J., Gytarsky, M, Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L., Miwa, K.,
Ngara, T., Tanabe, K. & Wagner, F. (eds.) 2003. Good Practice Guidance for Land Use,
Land-Use Change and Forestry. IPCC National Greenhouse Gas Inventories Program. The
Intergovernmental Panel on Climate Change (IPCC). Institute for Global Environmental
Strategies (IGES), Hayama, Japan. Available online at: http://www.ipcc-
nggip.iges.or.jp/public/gpglulucf/gpglulucf_contents.html; last accessed 12 August 2009.
Petersson, H. & Ståhl, G. 2006. Functions for below-ground biomass of Pinus sylvestris, Picea
abies, Betula pendula and Betula pubescens in Sweden. Scandinavian Journal of Forest
Research 21, Supplement 7: 84–93.
Pettersson, N. 1993. The effect of density after precommercial thinning on volume and structure in
Pinus sylvestris and Picea abies stands. Scandinavian Journal of Forest Research 8: 528–539.
Pretzsch, H. 2001. Modelliering des Waldwachstums. Parey, Berlin, 336 pp.
Rabinowitsch-Jokinen, R. & Vanha-Majamaa, I. 2010. Immediate effects of logging, mounding
and removal of logging residues and stumps on coarse woody debris in managed boreal
Norway spruce stands. Silva Fennica 44: 51–62.
Ranta, T. 2005. Logging residues from regeneration fellings for biofuel production – a GIS-based
availability analysis in Finland. Biomass and Bioenergy 28(2), 171–182.
Raulund-Rasmussen, K., Stupak, I., Clarke, N., Callesen, I., Helmisaari, H.-S. Karltun, E. &
Varnagiryte-Kabasinskiene, I. 2008. Effects of very intensive biomass harvesting on short and
long term site productivity. In: E. Röser, A. Asikainen, K. Raulund-Rasmussen and I. Stupak
(eds.). Sustainable use of forest biomass for energy – a synthesis with focus on the Nordic and
Baltic countries. Managing Forest Ecosystems Vol. 12, Springer.
Raymer, A.K.P., Gobakken, T., Solberg, B., Hoen, H.F. & Bergseng, E. 2009. A forest
optimisation model including carbon flows: Application to a forest in Norway. Forest
Ecology and Management 258: 579–589.
Repola, J. 2008. Biomass equations for birch in Finland. Silva Fennica 42(4): 605–624.
Repola, J. 2009. Biomass equations for Scots pine and Norway spruce in Finland. Silva Fennica
43(4): 625–647.
Repola, J., Ojansuu, R. & Kukkola, M. 2007. Biomass functions for Scots pine, Norway spruce
and birch in Finland. Working Papers of the Finnish Forest Research Institute 53. 28 p.
Rognstad, O. & Steinset, T.A. 2008. Landbruket i Norge 2007. Statistiske Analyser, 1–82.
87
�
Rudolphi, J. & Gustafsson, L. 2005. Effects of forest-fuel harvesting on the amount of deadwood
on clear-cuts. Scandinavian Journal of Forest Research 20: 235 – 242.
Rydberg, D. 2000. Initial sprouting, growth and mortality of European aspen and birch after
selective coppicing in central Sweden. Forest Ecology and Management 130(1-3): 27–35.
Rypdal, K., Bloch, V.V.H., Flugsrud, K., Gobakken, T., Hoem, B., Tomter, S.M. & Aalde, H.
2005. Emissions and removals of greenhouse gases from land, use, land-use change and
forestry in Norway. NIJOS report 11/2005
Rørstad, P.K, Trømborg, E., Bergseng, E. & Solberg, B. 2010. Combining GIS and forest
management optimisation in estimating regional supply of harvest residues in Norway.
Accepted Silva Fennica.
Röser, D., Asikainen, A., Raulund-Rasmussen, K. & Stupak, I. 2008. Sustainable use of forest
biomass for energy. A synthesis with focus on the Baltic and Nordic region. Managing Forest
Ecosystems, 12. Springer, 255 pp.
Saarsalmi, A. & Malkonen, E. 2001. Forest fertilization research in Finland: A literature review.
Scandinavian Journal of Forest Research, 16(6): 514–535.
Sexton, J.O., Bax, T., Siqueira, P., Swenson J.J. & Hensley, S. 2009. A comparison of lidar, radar,
and field measurements of canopy height in pine and hardwood forests of southeastern North
America. Forest Ecology and Management 257: 1136–1147.
Sikström, U, Jacobson, S., Johansson, U., Kukkola, M., Saarsalmi, A. & Holt-Hanssen, K. 2009.
Långtidseffekter på skogsproduktion efter askåterföring och kalkning - Preliminära resultat
från en pilotstudie. Rapport VÄRMEFORSK, Stockholm. April 2009. 22 pp. ISSN 1653-
1248.
Skovsgaard, J.P. & Vanclay, J.K. 2008. Forest site productivity: a review of the evolution of
dendrometric concepts for even-aged stands. Forestry 81: 12–31.
Snedecor, G.W. & Cochran, W.G. 1989. Statistical Methods. 8th Edition. Iowa State University
Press. 503 pp.
Solberg, S., Andreassen, K., Clarke, N., Torseth, K., Tveito, O.E., Strand, G.H. & Tomter S. 2004.
The possible influence of nitrogen and acid deposition on forest growth in Norway. Forest
Ecology and Management 192: 241–249.
Solberg, S., Næsset, E. & Bollandsås, O.M. 2006. Single-tree segmentation using airborne laser
scanner data in a structurally heterogeneous spruce forest. Photogrammetric Engineering &
Remote Sensing 72: 1369–1378.
Solberg, S., Astrup, R., Bollandsås, O.M., Næsset E. & Weydahl, D.J. 2010a. Deriving forest
monitoring variables from X-band InSAR SRTM height. In review.
88
�
Solberg, S., Astrup, R., Gobakken, T., Næsset, E. & Weydahl, D. 2010b. Estimating spruce and
pine biomass with interferometric X-band SAR. In review.
Spinelli, R., Cuchet, E. & Roux, P. 2007. A new feller-buncher for harvesting energy wood:
Results from a European test programme. Biomass and Bioenergy 31: 205–210.
Spinelli, R., Magagnotti, N. & Picchi, G. 2009. Complete tree harvesting as an alternative to
mulching in early thinnings. Forest Products Journal 59: 79–84.
Strand, L. 1967. Avsmalingstabeller for furu sønnafjells (Taper tables for Scots pine in Southern
Norway) Medd. Norske Skogf.ves. 22: 431–482.
Strand, L. & Braastad, H. 1967. Høydekurver for bjørk. Meddelelser fra det Norske
Skogforsøksvesen 22: 291–302.
Stupak, I., A. Asikainen, M. Jonsell, E. Karltun, A. Lunnan, D. Mizaraite, K. Pasanen, H. Pärn, K.
Raulund-Rasmussen, D. Röser, M. Schroeder, I. Varnagiryte, L. Vilkriste, I. Callesen, N.
Clarke, T. Gaitnieks, M. Ingerslev, M. Mandre, R. Ozolincius, A. Saarsalmi, K. Armolaitis,
H.-S. Helmisaari, A. Indriksons, L. Kairiukstis, K. Katzensteiner, M. Kukkola, K. Ots, H. P.
Ravn, & P. Tamminen. 2007. Sustainable utilisation of forest biomass for energy—
Possibilities and Problems: Policy, legislation, certification, and recommendations and
guidelines in the Nordic, Baltic, and other European countries. Biomass and Bioenergy 31:
666–684.
Ståhl, P.H. 2009. Produktionshöjande åtgärder. Skogsskötselserien, 16: 1–77.
Sverdrup-Thygeson, A. 2002. Key Habitats in the Norwegian Production Forest: A Case Study.
Scandinavian Journal of Forest Research 17(2): 166–178.
Sverdrup-Thygeson, A., Lie, M.H., Borg, P. & Bergsaker, E. 2002. Landskapsplan i skogbruket -
en veileder i landskapsplanlegging. NORSKOG og Prevista. 16 pp.
Swedish Forest Agency. 2008. Skogliga konsekvensanalyser 2008. SKA-VB 08. 157 pp.
http://www.skogsstyrelsen.se/episerver4/dokument/sks/aktuellt/press/2008/
Telenius, B.F. 1999. Stand growth of deciduous pioneer tree species on fertile agricultural land in
southern Sweden. Biomass and Bioenergy 16(1): 13–23.
Tham, Å. 1987. Gran och björk i blandning. Gallring i unga bestånd, forsöksuppläggning och
utgångsläge. (Summary: Mixed spruce and birch. Thinning in young stands, experimental
design and initial position.). Swedish University of Agricultural Sciences, Department of
Forest Yield Research, Report, 20: 1–79.
Tham, Å. 1988. Yield prediction after heavy thinning of birch in mixed stands of Norway spruce
(Picea abies ( L.) Karst.) and birch (Betula pendula Roth & Betula pubescens Ehrh.). Swedish
University of Agricultural Sciences, Department of Forest Yield Research, Report, 23: 1–36.
89
�
Tham, Å. 1989. Prediction of individual tree growth in managed stands of mixed Picea abies (L.)
Karst. & Betula pendula Roth & Betula pubescens Ehrh. Scandinavian Journal of Forest
Research 4: 491–512.
Tham, Å. 1994. Crop plan and yield predictions for Norway spruce (Picea abies (L.) Karst.) and
birch (Betula pendula Roth & Betula pubescens Ehrh.) mixtures. Studia Forestalia Suecica
195: 1–21.
Thies, M., Pfeifer, N., Winterhalder, D. & Gorte, B.G.H. 2004. Three-dimensional reconstruction
of stems for assessment of taper, sweep and lean based on laser scanning of standing trees.
Scandinavian Journal of Forest Research 19(6): 571–581.
Tomppo, E., Haakana, M., Katila, M. & Perasaari, J. 2008. Multi-source national forest inventory,
methods and applications. Managing Forest Ecosystems Vol. 18. Springer. 373 pp.
Trømborg, E., Bolkesjø, T.F. & Solberg, B. 2007a. Impacts of policy means for increased use of
forest-based bioenergy in Norway – a spatial partial equilibrium analysis. Energy Policy 35:
5980–5990.
Trømborg, E., Bolkesjø, T.F. & Solberg, B. 2007b. Skogbasert bioenergi til oppvarming –
økonomisk potensiale i Norge og effekt av økonomiske virkemidler. INA-fagrapport 9.
Institutt for naturforvaltning, Universitetet for miljø- og biovitenskap, Ås.
Tveite, B. 1977. Bonitetskurver for furu. Intern rapport.
Tveite, B. 1978. Bonitetskurver for gran. Medd. Nor. inst. skogforsk. 33: 1–84.
Valkonen, S. & Ruuska, J. 2003. Effect of Betula pendula admixture on tree growth and branch
diameter in young Pinus sylvestris stands in southern Finland. Scandinavian Journal of Forest
Research 18: 416–426.
Valkonen, S. & Valsta, L. 2001. Productivity and economics of mixed two-storied spruce and
birch stands in Southern Finland simulated with empirical models. Forest Ecology and
Management 140: 133–149.
Van der Zande, D., Hoet, W., Jonckheere, I., van Aardt, J. & Coppin, P. 2006. Influence of
measurement set-up of ground-based LiDAR for derivation of tree structure. Agricultural and
Forest Meteorology 141: 147.
Vestjordet, E. 1967. Funksjoner og tabeller for kubering av stående gran. Functions and Tables for
Volum of Standing Trees. Norway Spruce. Meddelelser fra Det Norske Skogforsøksvesen 22:
539–574.
Vestjordet, E. 1977. Precommercial thinning of young stands of Scots pine and Norway spruce.
Meddelelser Fra Norsk Institutt for Skogforskning, 33(9): 311–436.
90
�
Walle, I.V., Van Camp, N., Van de Casteele, L., Verheyen, K. & Lemeur, R. 2007a. Short-rotation
forestry of birch, maple, poplar and willow in Flanders (Belgium) I - Biomass production
after 4 years of tree growth. Biomass and Bioenergy 31: 267–275.
Walle, I.V., Van Camp, N., Van de Casteele, L., Verheyen, K. & Lemeur, R. 2007b. Short-
rotation forestry of birch, maple, poplar and willow in Flanders (Belgium) II. Energy
production and CO2 emission reduction potential. Biomass and Bioenergy 31: 276–283.
Watt, P.J. & Donoghue, D.N.M. 2005. Measuring forest structure with terrestrial laser scanning.
International Journal of Remote Sensing 26: 1437–1446.
Weydahl, D.J., Sagstuen, J., Dick Ø.B. & Rønning, H. 2007. SRTM DEM accuracy assessment
over vegetated areas in Norway. International Journal of Remote Sensing 28: 3513–3527.��
Wilhelmsen, G. & Vestgjordet, E. 1974. Preliminary dry wood weight tables for merchantable
stems and stands of Norway spruce in Norway. Reports of the Norwegian Forest Institute
31(5): 183–240.
Wilhelmsson, L. 2001. Characterisation of wood properties for improved utilisation of Norway
Spruce an Scots pine. Doctoral thesis. Swedish University of agricultural sciences. Uppsala,
Sweden. 110 pp.
Zeide, B. 2001. Thinning and growth: A full turnaround. Journal of Forestry 99: 20–25.
Zianis, D., Muukkonen, P., Mäkipää, R. & Mencuccini, M. 2005. Biomass and Stem Volume
Equations for Tree Species in Europe. Silva Fennica Monographs (4): 1–63.
Øhman, K. & Eriksson, L.O. 1998. The core area concept in forming contiguous areas for long-
term forest planning. Canadian Journal of Forest Research 28: 1032–1039.
Øyen, B.-H., Øen, S. & Skatter, J. 2001. Planteavstandens betydning for bestandsutvikling og
lønnsomhet i en vestnorsk granplanting. Rapport fra skogforskningen 8/01: 1–19.
Åström, M., Dynesius, M., Hylander, K. & Nilsson, C. 2005. Effects of slash harvest on 621
bryophytes and vascular plants in southern boreal forest clear-cuts. Journal of Applied
Ecology 622: 1194–1202.
91